JABE-ARC-07 GVLU /?/V/ 7LS ) STUDY AND EVALUATION OF FERRO-CEMENT FOR USE IN WIND TUNNEL CONSTRUCTION prepared under Contract No. NAS2-5889 for NATIONAL AERONAUTICS AND SPACE ADMINISTRATION Ames Research Center Moffett Field, California (NASA-CR-114501) STUDY AND EVALUATION OF FERRO-CEMENT FOR USE IN WIND TUNNEL CONSTRUCTION H.J. Larsen, Jr. (Blume (John A.) and Associates Research) Jul. 1972 155 p CSCL 13C G3/32 July 1972 N72-33916 Unclas 44899 . _ John A. Blume & Associates, Engineers San Francisco, California Reproduced by NATIONAL TECHNICAL INFORMATION SERVICE U S.Department of Commerce Springfield VA 22151 , .__
Transcript
JABE-ARC-07
GVLU /?/V/ 7LS )
STUDY AND EVALUATION
OF
FERRO-CEMENT
FOR USE IN
WIND TUNNEL CONSTRUCTION
prepared under
Contract No. NAS2-5889
for
NATIONAL AERONAUTICS AND
SPACE ADMINISTRATION
Ames Research Center
Moffett Field, California
(NASA-CR-114501) STUDY AND EVALUATION OFFERRO-CEMENT FOR USE IN WIND TUNNELCONSTRUCTION H.J. Larsen, Jr. (Blume (JohnA.) and Associates Research) Jul. 1972155 p CSCL 13C G3/32
July 1972
N72-33916
Unclas44899
. _
John A. Blume & Associates, Engineers
San Francisco, California
Reproduced by
NATIONAL TECHNICALINFORMATION SERVICE
U S.Department of CommerceSpringfield VA 22151
, .__
FOREWARD
The successful completion of a state-of-the-art survey such as the
one reported herein is dependent on the cooperation of a number of
individuals. Professor R. B. Williamson, University of California,
Berkeley was especially helpful as was Professor S. P. Shah, University
of Illinois at Chicago Circle. The assistance of those in industry
such as M. E. Irons of Fibersteel Corporation and D. A. Seymour, Naval
Architect is also appreciated. Professor W. J. Venuti, California
State University, San Jose helped develop the laboratory test criteria
and supervised the testing of all samples.
The work in the Blume office was conducted under the general super-
vision of Roland L. Sharpe, Principal-in-charge, and James E. Boyd,
Project Manager. Henry J. Larsenr Jr.-was the principal investigator
responsible for compilation, review and evaluation of the data as
well as planning and conduct of the laboratory testing program.
JOHN A. BLUME & ASSOCIATES. ENGINEERS
JABE-ARC-07
STUDY AND EVALUATION
OF
FERRO-CEMENT
FOR USE IN
WIND TUNNEL CONSTRUCTION
CONTENTS
Page
FOREWARD ---------------------------------------- ii
I. INTRODUCTION ------------------------------------- ----
II. THE NATURE OF FERRO-CEMENT -------------------------------- 4
A. General Properties ------------------------------------ 4
B. History ---------------------------------------- 5
C. Uses of Ferro-cement ---------------------------------- 7
III. THE STATE-OF-THE-ART IN FERRO-CEMENT ------------------- 9
A. Strength Properties ----------------------------------- 9
be based on specific test programs to establish an optimum design.
Principal areas for further work include more comprehensive fatigue
testing, reinforcement and mix design studies, durability studies,
and full-scale load testing.
1.
- 3-
JOHN A. BLUME & ASSOCIATES, ENGINEERS
II. THE NATURE OF FERRO-CEMENT
The ferro-cement concept is as old as that of reinforced concrete, but
its use as a structural material has received widespread attention only
in the last several decades. Ferro-cement is given a brief description
and then the history of its development and current and proposed appli-
cations are summarized in the following text.
A. GENERAL PROPERTIES
Ferro-cement basically consists of a thin-shell of Portland cement
mortar heavily reinforced with steel wire. The reinforcement generally
consists of several layers of light gage steel wire mesh. Typical
shell thicknesses are from 3/8 inch to 1-1/2 inches. Sometimes steel
reinforcing bars are sandwiched between the layers of wire mesh.
Figure 1 shows two common types of ferro-cement reinforcing. Section
la of Figure 1 shows a network of steel reinforcing bars overlain with
layers of wire mesh that is impregnated with mortar. The use of rein-
forcing bars with wire mesh adds to the strength of the material and
also provides a means of establishing the structural shape. Section lb
of Figure 1 is reinforced only with wire mesh and the required shape is
obtained through external means such as casting molds. Because of the
close spacing of reinforcement evident in Figure 1, care has to be
taken in placement of mortar.
The material constituents of ferro-cement and those of more commonly
recognized reinforced concrete are very similar, although the propor-
tions of the materials used in ferro-cement give it several different
and unique properties. The behavior of ferro-cement during strain and
cracking demonstrates a synergistic effect. Because the steel reinforcing
4i
JOHN A. BLUME & ASSOCIATES. ENGINEERS
57EEL REG5AR'.
W/IRE MEfH
JEC TI ON
V/RE MESH). .~-* p. or.
- D g ' . 0 * / .**3 _ .w '. *.IB S B A* *.(- @~9 *. ** . l e . , ,,_
Se CTI ON
c . - 12
JFIURI I - TYP/CAL FRRO-CE41ENT7 SJCT/O/0J"JOHN A. BLUME & ASSOCIATES. ENGINEERS
is very evenly dispersed through the cross section, formation of cracks
in the mortar is inhibited and the stress level at the onset of tensile
cracking is significantly increased over plain mortar and ordinary re-
inforced concrete. The surface area of reinforcement is the parameter
most-closely related to the strain and cracking behavior of ferro-cement.
Compared to ordinary reinforced concrete, ferro-cement exhibits a
uniquely high surface area of reinforcement relative to its total volume.
This is one of the characteristics that differentiate ferro-cement from
reinforced concrete.
Much research in the last ten years has been done on a material closely
related to ferro-cement, commonly referred to as "fiber reinforced
concrete." This material consists of a Portland cement concrete or mor-
tar reinforced with short, small-diameter wires. A typical wire size is
1 inch long by 0.02 inch in diameter. Chief advantages of fiber rein-
forced concrete are good dispersal of reinforcement and cost reductions
resulting from the fact that the short fibers can be mixed and placed
with the concrete or mortar matrix, using the same equipment. Because
of the short fiber length, however, the principle mode of failure for
this material is bond failure and wire pull-out. This type of failure
is quite sudden and is undesirable in concrete structures. To date the
primary applications of fiber reinforced concrete have been experimental
concrete highway and airport runway slabs ranging in thickness from 4 to
6 inches. It has also been used for some thin shells. Relative to
ferro-cement, however, the current state-of-the-art in fiber reinforced
concrete does not warrant consideration at this time for use in the
construction of wind tunnel shells. A bibliography of references on
fiber reinforced concrete work, most of which were reviewed during this
study, is included in Chapter IX.
B. HISTORY
The development of ferro-cement as a structural material has centered
around its applications for the construction of marine-craft. One of
-13'5 -
JOHN A. BLUME & ASSOCIATES. ENGINEERS
the first uses of ferro-cement and of reinforced concrete of any type
was the construction of several small boats in France by Lambot in
1855.1 But the modern development of ferro-cement began in the early
1940's with the Italian engineer Pier Luigi Nervi. He designed and
built a number of sailboat and motorboat hulls, as well as some
architectural structures, using thin-shell ferro-cement construction.
His investigations into the properties of ferro-cement can be summarized
in the following statement:
The fundamental idea behind this new reinforcedconcrete material ferro-cement is the well knownelementary fact that concrete can stand largestrains in the neighborhood of the reinforcementand that the magnitude of stress depends on thedistribution and subdivision of the reinforcementthroughout the mass of concrete.2
Since the 1940's the majority of ferro-cement construction has been
amateur-built, "backyard" boats ranging up to 60 feet in length. Poor
results from some of these early projects caused ferro-cement to fall
into some disrepute as a legitimate structural engineering material.
Within the last decade, however, its potential has been recognized by
a number of serious commercial boatbuilders as well as public and
private research institutions around the world. Commercially built
fleets of sailboats, power vessels, and cargo barges now exist or are
planned in the United States, Canada, the United Kingdom, Australia,
New Zealand, the Soviet Union, and China. Basic research on the
engineering properties of ferro-cement has been carried out in most
of these countries. A large amount of the research work done to date
has been valuable in determining the engineering properties of the
material; more work is now in progress, and more is needed to establish
ferro-cement as a viable engineering material for general structural
use. In Chapter III of this report the current state of knowledge of
ferro-cement is summarized and areas requiring additional research are
pointed out.
.14- 6-
JOHN A. BLUME & ASSOCIATES, ENGINEERS
C. USES OF FERRO-CEMENT
As stated in the preceding section, the primary focus in ferro-cement
development and construction has been in marine craft hulls. The
material is especially suited to this type of construction because it
can be molded or formed into virtually any shape in a monolithic unit.
It also has relatively high rigidity, relatively high compressive and
flexural strength and resistance to cracking, and is low in material
cost relative to other boat-building materials. Also, it is highly
resistant to fire and most corrosive elements and is easily repairable.
Marine craft that have been built of ferro-cement include private
sailing and motor yachts from 30 feet to 60 feet in length, as well
as commercial fishing and cargo vessels up to 180 feet long. The
Naval Civil Engineering Laboratory at Port Hueneme, California, has
studied ferro-cement for prefabricated construction panels.3 The
Canadian government is currently sponsoring basic research and proto-
type construction of ferro-cement cargo barges.4 The United States
Navy Naval Ship Research and Development Center is presently engaged
in research and prototype construction of 24-foot, high-speed motor
launches with ferro-cement hulls as thin as 3/8 inch.5 Ferro-cement
has also been used extensively for the construction of marina floats.
Pier Luigi Nervi also pioneered the use of ferro-cement for buildings
and other civil engineering structures. He used ferro-cement in
applications such as walls for small buildings and precast units for
stadium roofs. Ferro-cement has more recently been used for lining
mine shafts and tunnels in Eastern Europe and for-decorative paneling
in Australia.6 Its use has been proposed for many types of tanks,
including liquid natural gas containers.7
Architectural and civil engineering applications of ferro-cement have
not kept pace with marine applications since Nervi's first use of the
'7 -
JOHN A. BLUME & ASSOCIATES. ENGINEERS
material. There appear, however, to be definite cost advantages to
ferro-cement in certain kinds of applications. It is a material that
can be engineered to a high degree of precision, yet can be constructed
by semi-skilled or unskilled labor, using relatively inexpensive
materials. As more is learned about the engineering properties of
ferro-cement, these advantages should lead to wider usage for many
types of civil engineering structures as well as marine structures.
A major ferro-cement study is currently being sponsored by the National
Academy of Science.8 The purpose of this study is to establish the
engineering properties of ferro-cement sufficiently well that its
low material costs and labor-intensive fabrication can be utilized
by developing countries for applications such as marine craft and grain
storage structures.
-16
JOHN A. BLUME & ASSOCIATES, ENGINEERS
III. THE STATE-OF-THE-ARTIN FERRO-CEMENT
The state-of-the-art pertaining to important ferro-cement characteris-
tics including current information on the engineering properties of
the material and the various techniques by which ferro-cement structures
are being fabricated is presented in this chapter.
Most ferro-cement work to date has been related to marine construction.
Applications in other fields will undoubtedly give rise to questions
about its material properties and fabrication methods. Recommendations
are made where the necessity for additional work on material properties
or fabrication techniques is indicated, especially in relation to its
use for wind tunnel construction. These recommendations are summarized
in Chapter VIII of this report.
The important material properties of ferro-cement cover a wide range
of engineering design parameters. The following discussions are
based on review and analysis of current ferro-cement research and
construction practice as well as consultation with individuals active
in these fields. Building codes or standardized design procedures
have not been established for ferro-cement. The following sections,
therefore, present quantitative information as well as insight into
the characteristics and behavior of ferro-cement so that appropriate,
economical design methods can be developed for specific structural
applications.
A. STRENGTH PROPERTIES
The behavior and capacity of ferro-cement subjected to various kinds
of static load as well as fatigue and impact loads are discussed in
the following sections.
JOHN A. 9-LUME ASSOCIATES, ENGINEERSJOHN A. BLUME &e ASSOCIATES, ENGINEERS
1. Compressive Strength
Ferro-cement compressive strength is primarily dependent on the
compressive strength of the mortar matrix. Typical ultimate values
are 5,000 to 10,000 psi at 28 days. Based on work at the
Massachusetts Institute of Technology in 1969, J. F. Collins and
J. S. Claman9 reported that the inclusion of wire mesh reinforcement
does not significantly increase or decrease the ultimate mortar com-
pressive strength. Based on the requirements of the American Concrete
Institute Building Code (ACI 318-63) Requirements for Working Stress
Design of Concrete, ferro-cement working stresses should be limited
to 25 percent of ultimate in uniaxial compression and 45 percent of
ultimate in flexural compression.
2. Tensile Strength
The behavior of ferro-cement in tension represents a significant
departure from that of ordinary reinforced concrete. As a result
of the high degree of dispersion of reinforcement, the first tension
cracks in the mortar matrix form at stress levels significantly higher
than for unreinforced mortar or ordinary reinforced concrete. In
addition, crack spacing is generally close and crack width is small.
The dispersal of small diameter reinforcement in the mortar results
in a material that exhibits a relatively homogeneous behavior during
strain and cracking.
The two most significant points of interest in the tensile behavior of
ferro-cement are the stress at formation of the first crack in the mortar
and the ultimate strength. Figure 2 shows the load-elongation relation-
ship for a typical tensile test reported in 1970 by S. P. Shah of the
Department of Materials Engineering, University of Illinois at Chicago
Circle.7 The material undergoes elastic elongation prior to first cracking,
which is a distinct point in the material behavior. After first cracking,
the behavior is quasi-elastic with a reduced modulus.
18- 10 -
JOHN A. BLUME & ASSOCIATES. ENGINEERS
STJRA IN
2 - TY,/CA L a04 -FI GuiZE ELONG ACR eoM
7T ON 7rE)A V1 O )RCF. 7 ) -_19
JOHN A. BLUME & ASSOCIATES, ENGINEERS
/o00
LIP
()
I-
e.
r-
i0
'UU~ltov~
50
0o
The formation of a tensile crack through the ferro-cement cross
section is the result of an accumulation of micro-cracks in the
material, which begin forming at low load levels and increase in
numbers as the load is increased. The presence of micro-cracks is
typical of cementitious materials. In ferro-cement, however, the
wide dispersion of reinforcement throughout the mortar matrix restricts
the propogation of these micro-cracks to the vicinity of an individual
wire. The formation of the first observable crack indicated in
Figure 2 is the point where the micro-crack width becomes large
enough that a crack propogates through the entire section.
As tensile stress is increased following the first crack, additional
cracks occur. Both formation and widening of cracks are related to
localized bond failure in the vicinity of the micro-cracks, but the
wide dispersion of reinforcement restricts the extent of bond failure.
This results in a larger number of cracks of small individual width.
The stages of loading and cracking of ferro-cement have been reported
in more detail based on recent work in Poland by J. R. Walkus and
T. G. Kowalski.6 The behavior of ferro-cement in tension (and
similarly in flexure) is summarized in Figure 3 and in the following
statement from their work:
In the initial stage the material behaves in a linearly-elastic manner when loaded. Elastic deformations occurat this stage in both metal [wire mesh reinforcement] andcrystalline grids [hydrated cement crystals] as well asin colloids [unhydrated materials].
With a further increase in stress, ferro-cement becomesquasi-elastic. The relatively small plastic strains ofthe colloids are restrained by the elastic deformation ofthe metal wires...The micro-cracks are invisible to thenaked eye and are difficult to observe even when opticalinstruments are used. When the load is released, evenoptical instruments will not enable the positions wherethe micro-cracks have occurred to be detected -- suchis the extent of their closing-up.
JOHN A. BLUME & ASSOCIATES, ENGINEERS
U)v)tu11
I..
uI.
%I-)IT
(20) (50) (oo00)S RA /IN
(CRACK WIDTHS IN b/RACKETS, ,1MCRONS'I)
FIGURE 3a
CRACKSTAG E /MATERIAL S7ATre W1/I4TH, TECHNJOLOGICAL SrANT
MICRONS
. LINEARLY- ELASTIC 20 COA41LETE
QUASI - ELASTIC .50 NOCORR O Ve I2Z
ELASTO - PLASTIC /00 H0NCORR/ v.6
IF PLAS TIC >100 CORROsI VE
FIG IGR L 3b
F/GURE 3,- WORKING STAGES OF FERRO-CEAM4lT(FROM RF=. 6) . -. z
JOHN A. BLUME & ASSOCIATES, ENGINEERS
IIIIIIIII
!III III'IIIIIIII
IIIII
I
I
II
II
These two stages -- the linearly-elastic and the quasi-elastic -- constitute the practical elastic workingrange of ferro-cement. A further increase in stresscaused very definite plastic deformation of the colloidsas well as crystalline grids, which is in turn resistedby the metallic grids of the reinforcement. This is thetime of the formation and widening of exploitationalcracks [i.e., cracks caused by loading]...
After the 50-micron limit has been reached the processof crack widening continues at a uniform rate. Coopera-tion between concrete and steel continues up to theattainment of a crack width of 100 microns and thereafterthe steel alone carries all the tensile forces.6
The Technological State shown in Figure 3b refers to the dependence of
ferro-cement corrosion resistance on crack width. This is discussed
further in the section on corrosion resistance.
The first crack strength of ferro-cement in tension has been found to
be related to volume percentage of reinforcement and surface area of
reinforcement. The most significant of these two parameters has been
found to be surface area of reinforcement.6,7,10' 11 The most commonly
used measure of this quantity is specific surface of reinforcement
(Sp), which is defined as the surface area of the reinforcement (in
direction of load) divided by the total volume of ferro-cement. For
the same volume of wire reinforcement, a large number of small
diameter wires has a high value of specific surface while a small
number of large diameter wires has a lower specific surface value.
Figure 4, from experimental studies by S. P. Shah,7 shows an increase
in the tensile cracking stress of the ferro-cement related to an
increase in the steel content. Figure 5a, also from Shah's work,
shows a strong relationship between stress at first crack and specific
surface of reinforcement. The tests represented in Figure 5a show
first cracking stress is increased by a factor of about three over
unreinforced mortar for high values of specific surface. This result
demonstrates a synergistic effect in ferro-cement wherein the presence
-'12 -
JOHN A. BLUME & ASSOCIATES, ENGINEERS
23
/000
In
LiOk1)u'
doo
GOO
u .
O L0
I I
A
C
I I I I I.0 2..0 3.0 4.0
VOCUM PE/RCENT OF RE/FoRCe446Nr/N LOA.D 10/RCCT/ON
p/sGURe 4- TENJILe JrRe.Sr AT FIR7T CCRACKVOLUM E o,= RgeN.loRc E4e N7CFRo / REP. 7 )
5.0
VJ.
,JOHN A. BLUME & ASSOCIATES. ENGINEERS
I I
24
,90
0---- -o
0
0o
_ wo °6)
0
I I , I . I
e 3 4,PEC/F/C L/R SUA CE OF AR/MFORCEMA4/EM
/AN OADOIAG D/ReCT 7ON,T/Ne///N 3
3
2
i
00
F/GURE 6 6
0 " O &
I I I II I1
SPECi/FC SURFACE OF REI'FORC/EMEN1T/NI LOADFIG D/R EC7O'10,1AII//I 3
of wire mesh improves the tensile cracking behavior of the mortar
matrix. Figure 5b shows a relationship between decreased crack
spacing (hence decreased crack size) and increased specific surface
of reinforcement.
Although the first cracking strength of ferro-cement is improved by
the use of wire reinforcement, the first cracking strength is usually
well below the ultimate capacity. For many structural applications
of ferro-cement, the presence of cracks of a limited size may not be
detrimental. Shah has proposed a ferro-cement design procedure based
on a maximum specified allowable crack width under service loads.1 2
For example, allowable crack widths would be quite low for structures
that must be watertight or resistant to corrosive elements, and
relatively higher for structures in a less harsh environment. Shah
is currently conducting studies to establish a relationship between
crack width, specific surface of reinforcement, and reinforcing steel
stress for ferro-cement in tension. A hypothetical set of design
curves based on this kind of research is shown in Figure 6. Specifi-
cation of the maximum allowable crack width and selection of a value
of a specific surface would give an allowable design stress.
No direct relationship has been found between the tensile cracking
behavior of ferro-cement and the properties of the mortar matrix.
During ferro-cement testing at MIT in 1969, however, J. F. Collins13
observed that poor bonding between mortar and steel mesh leads to a
lower tensile cracking strength. Poor bond in the wires perpendicular
to the direction of stress creates the equivalent of a void in the
region of the wire. This leads to stress concentrations and premature
cracks perpendicular to the direction of stress. Based on this
finding, care should be taken in mix design and preparation of rein-
forcement to insure proper bond.
The ultimate tensile strength of ferro-cement has been found by Shah7
to be dependent solely on the tensile capacity of the reinforcement.
This is clearly demonstrated by the test results shown in Figure 7.
- 13 -JOHN A. BLUME & ASSOCIATES, ENGINEERS
-/15 KI STSELSTRESS
JPEC//l/C SURFACE OF E&/NP9ORC6M6Y T
F/GURee G - IYPOTHRTIE/CA.L etO- CEMEN'T EZ5/GPN'CULVE/S V AJE5G ON CIA CK W/7 Tt4
JOHN A. BLUME & ASSOCIATES, ENGINEERS
26
I-
Q
Uv3"Itok
/x/
p
9 //9/
LI/NE OF EQA4.L/TY(CLNE oF POlJ 7*T A LONG
WLNICH OIZO/NA4T ANOA /Sc/JS ,4ARE 60A4 .L)
f I I I I I
4 5LoAin CAR,'YIN CAPA4CI/7Y OF ,R?/vFO,/CEE/V7r
IN LCo4/OlA G //RECr"/ON, /6 x /O0
V..
JOHN A. BLUME & ASSOCIATES. ENGINEERS
27
5
4
13
en
0o
3(U
0
Q
*K
0
'Ii
I.-
/d
I_- P'p,
- P
/I , I
2 3
7- ULt7-IMATe 4.OA) OF COMP/OSI7ULLTIMA 7T CAPA C / TY OF RleNFOfeCaEV1N T(CFRoM REF. 7 )
C 60% ·
3. Flexural Strength
Ferro-cement in bending exhibits unique properties much like those
discussed for tension. A relationship between increasing stress at
first crack and increased volume and specific surface of reinforcement
has been observed, and at ultimate flexural capacity its strength is
related primarily to the capacity of the reinforcement.
Typical load-deflection curves from flexural tests conducted in 1970
by J. E. Tancreto and H. H. Haynes of the Naval Civil Engineering
Laboratory, Port Hueneme, California,14 are shown in Figure 8.
Similar to tension, flexure behavior exhibits an initial linear de-
flection followed by cracking of the mortar and further linear
deflection prior to yield and ultimate failure. The observations of
Walkus and Kowalski6 which are summarized in Figure 3, also apply to
the flexural behavior of ferro-cement. Based on their findings the
elastic behavior is limited by the formation of the first loading
crack. The first visible crack observed by Tancreto and Haynes
(Figure 8), however, is probably at a higher stress level than the
elastic limit proposed by Walkus and Kowalski.
Figure 9, which contains data obtained by Tancreto and Haynes, shows
a relationship between increased specific surface of reinforcement
and increased flexural stress at first crack. The formation of
flexural tension cracks results from the same behavior as discussed
for pure tension cracks. These cracks reduce the flexural rigidity
of the cross section. Its effect is observed as a reduction in the
modulus of elasticity for a flexural load test. This behavior is
verified in work done at MIT15 and the University of Michigan.16
The concept of a design procedure based on crack width as related
to stress and specific surface being studied by S. P. Shah 12 can
also be applied to ferro-cement in flexure. A design criterion
-- 14
JOHN A. BLUME 8c ASSOCIATES, ENGINEERS
. ;29
0./0 O. 20 0.30 o0.o os.c
O6FLeC.ET/ON, IN.
F/eIURE 8- TYPI/CAL FLEKUIA CO OA4o--DEFC'ECT/ONA-HAV/OR£ (,ROM ReF. /4)
ICeHN A RI IIMF P, AARSOCIATES, ENGINEERS
- 30
_ 00
.40
I/ *
.
* 0."
. .0, * *./
// '
lI
4 G 8 /0P/CIP/C J4URAACE OF REINFoRCCM/N7N
IN L04, iOIReeCT/ON,'-/N2//N
FIGURe 9 - FLEXURAL JTRfS$ AT F/eS7 CRACK ?Z.
JOHN A. BLUME & ASSOCIATES, ENGINEERS
2000o-
Iu'
i-A)
U U1%
/5001-
/000
/
JEC/o/C JUeFACg ,0, R.//FO)CESCTCiFReoM RcF, 14)
I 1 I I/a~o0
31
based on maximum allowable crack width under service loads for a given
structural application would then permit selection of an optimum re-
inforcement configuration and corresponding allowable design stress.
The current work by Shah is being done for pure tensile stress only.
Additional work would be required to establish design curves of this
type for ferro-cement in flexure.
An experimental method for studying concrete is currently being used
at the University of California at Berkeley whereby very thin sections
are cut from specimens after loading and studied microscopically.8
Such a method could be used to observe the formation and width of cracks
and also the elastic and inelastic deformation behavior of mortar and
reinforcement.
Unlike its behavior at cracking load, the behavior of ferro-cement at
ultimate flexural capacity demonstrates a primary dependence only on
the ultimate load carrying capacity of the steel reinforcement. This
result corresponds to similar findings mentioned above for the tensile
strength of ferro-cement. Tancreto and Haynes14 have found that the
ultimate flexural capacity of ferro-cement can be predicted quite
accurately by computing the capacity of a cross section consisting
only of the reinforcing steel using the following equation:
Mult fult c)
where
Mu = Ultimate flexural moment, in.-lb.
ful = Ultimate tensile strength of wire reinforcement, psi
I = Moment of inertia of wires about neutral axis, in.4
c = Distance from neutral axis to extreme tension wire, in.
- 15 -
JOHN A. BLUME & ASSOCIATES, ENGINEERS
The use of steel reinforcing bars in addition to wire mesh (Figure 1),
which increases the volume percentage of reinforcement, gives higher
reinforcement capacity and hence higher ferro-cement ultimate moment.
4. Shear Strength
There is a limited amount of shear strength data available with respect
to design of ferro-cement for shear loads. Available data indicate
that shear strength is dependent on the volume of reinforcement for
shear loading normal to the plane of the material.
Results of some of the earliest ferro-cement research work, done in
1959 in Ireland by L. D. G. Collen and R. W. Kirwin,17 given in
Figure 10, show increased shear strength is related to increased
steel content. These results are substantiated by Claman's1 5 observa-
tion that the inclusion of reinforcing bars in the cross section,
which generally increases steel content, results in increased shear
capacity.
In his work at MIT, Claman1 5 observed that the yielding of ferro-cement
in shear under loading normal to the plane of the material is accom-
panied by slippage of the reinforcement and abrupt change in the slope
of the load-deflection curve. This kind of behavior is undesirable
and should be considered to represent the ultimate shear load of
ferro-cement. To prevent this kind of yielding, design shear stress
should be kept well below this level, similar to the design of rein-
forced concrete members without shear reinforcement. Although design
shear stress will be relatively low, flexural or tensile stresses are
generally more critical for thin-shell structures.
5. Modulus of Elasticity
Studies of the modulus of elasticity of ferro-cement show a relation-
ship between increased volume percentage of reinforcement in the
- 16 -JOHN A. BLUME & ASSOCIATES, ENGINEERS
33
/5
cu
W-.
0
I/ -
0
010-
C4:1 I I I I
SO f2.5 . 3,0 S,5VOLUME PEeCCNT OF R~E/VFOCEME/NT7
/N ONE I//ECCT/ON
FILURe tO - U LTIM4A TE SHEA J TIZESVJ'. VOLUME OF RE / NOtCEM EN7·F o2M RE-F. /7)
JOHN A. pI I14'F P. AC " 'I'TFC Fhlr:-I, '" -
34
direction of loading and increased modulus. A distinct decrease in
modulus after formation of first crack has been observed in most
studies.
The load-elongation curve for a ferro-cement tensile test shown in
Figure 2 shows a decrease in modulus of elasticity after the first
crack. Similar two-phase behavior for flexure tests has been found
by Tancreto and Haynes (Figure 8) but the change in slope is generally
less abrupt. Shah has found that an estimate of the modulus of
elasticity in tension can be obtained from the law of mixture of
composite materials as follows:7
Before first crack: E = EM
+ ERLVL
After first crack: E = ERLV L
where
E = Modulus of elasticity of ferro-cement
EM = Modulus of elasticity of mortar
ERL = Modulus of elasticity of mesh in load direction
VL = Volume fraction of reinforcement in load direction
The test results by Shah for modulus of elasticity of ferro-cement in
tension are shown in Figure 11. It can be seen that modulus of elastic-
ity is directly related to volume of reinforcement. This was also
observed in early research in Ireland by Collen.17
The modulus of elasticity of ferro-cement in flexure or compression has
not been studied in detail; however, a dependence on mortar strength and
-17 -
JOHN A. BLUME & ASSOCIATES. ENGINEERS
0.5 1.0 /.5
PERCE/lT VOLUME OF REI/FORCEMS4NT7 VLLOADING D)IlRECT/ON
0-- CVU O / C4 7
PERCENT VOLUME OF R/tVFORC6EENT, V,LOA/DING D/ARECTIO/,
FIGURE I1- COMP/517 MA40DUC./US OF
0.5 /.0
JOHN A. BLUME & ASSOCIATES, ENGINEERS
.. 35
1.0
O.B0,6
0.4
o0
)7VO,
QC'Utjj
4' k u
>, -Z -4
11
"I1-4i
0
0
3
Q0
4'
LI
Q
UO
/.0
0.8U
k 0.2
o.
0
000
0I I
/.5I I I
Z.5
c ELAS7r/C/TY / r7EAs/o
F,:ROM REFP 7)
/.2
E -RL V,
36
volume of reinforcement, similar to tension would be expected. The
load-deflection results of Tancreto and Haynes 14 indicate values for
flexure of about 4,000,000 psi before cracking and about 1,000,000 psi
after cracking.
6. Fatigue Resistance
Ferro-cement behavior under repeated loading has been largely ignored
in published research work, although recent unpublished testing work
in the Naval Ships Research and Development Center at Annapolis in-
cluded some fatigue study. It is expected that fatigue resistance of
ferro-cement will be dependent on volume and specific surface of re-
inforcement, similar to the behavior of ferro-cement under tensile and
flexural static loading, and also on the range and magnitude of cyclic
loading.
Some early fatigue test results were obtained in about 1963 by an
English builder of ferro-cement boats, Windboats, Ltd.1 8 Results are
for bending fatigue of four samples 22 inches by 5 inches by 0.65-inch
thick, and are as follows:
More recent work completed in 1971 by the United States Naval Ships
Research and Development Center1 9 is shown in Figure 12. These data
show test results for samples with 6.5 percent reinforcement,
C - 18-JOHN A. BLUME & ASSOCIATES, ENGINEERS
Nominal StressSample Levels, ps Cycles Remarks
Levels, psi
A +625 2,000,000 cracked-544
B +700 2,000,000 no fracture-600
C t1100 100,000 cracked
D t1185 100,000 cracked
200oo
.N
Iq
J
*U
N
0j(J
4000
3800
J400
3200
3000
e0oo/C /o0 /0O j0'
NUA1MER, OF CYC. ES
.- + )CNO7EJr JA/MtoLE/W/D AO'7 FA/L
F /G URE. .2.- _fERI. - CEMAN .- f TR/U.E 'ES S(PFROM RFi. i )
JOHN A. BLUME & ASSOCIATES, ENGINEERS
\TLORET7CAL dURVE
.________ R A:-TRElED.Q .T0AT/64e.
.J7eNe67W OF PV/RE
,=e/CcMeN7,T1*
:)3
38
measuring from 3-to 6-inches wide and constant thickness, loaded as
cantilever beams with a 6-inch constant strain zone. Loading range
was from positive maximum load to negative maximum load and the
frequency was from 6 -to 18-cycles per minute. The datum points represent
the point when "fatigue cracks were noted or when the specimen had
accumulated 2,000,000 cycles." The theoretical curve indicated is based
on fatigue data for the wire reinforcement and application of the theory
of transformed sections. Thus it is assumed this curve is based on
fatigue failure as determined by fatigue yielding of the reinforcement.
It is important to note that the loading ranges used were well above
the load to cause first cracking of the mortar. The report concerning
this work concluded that "based on verification of the theoretical curve
by the data it appears that the endurance limit (fatigue yielding) for
this ferro-cement composition is about 3,000 psi." As seen in Figure
12 this endurance limit is for a maximum of 2-million loading cycles.
Static flexural tests reported in this same document indicate that the
static stress to produce reinforcement yielding is about 6,200 psi.
Thus the fatigue limit of 3,000 psi appears to represent about 50
percent of the static yield stress.
The above test results provide important and useful fatigue information.
For design of important or unusual structures, however, much additional
information is needed. For example, data on higher frequency loading,
various loading ranges, and higher number of load cycles would be
significant. A study of the long term effect of strength gain in
cementitious materials on fatigue behavior of ferro-cement would also
be significant. Whereas the Naval Ships Research and Development Center
data (Figure 12) give useful information in the range of ferro-cement
yield load, additional work is necessary to determine the effect of
fatigue on the first crack load. This information would be important
for structures where cracking is to be avoided. In addition, the
previously described work by Shah in establishing crack width as a
design parameter could be adapted to fatigue loading. Design curves
would relate crack width to stress level, specific surface, and also
number of fatigue loading cycles.
- 19 -
JOHN A. BLUME & ASSOCIATES. ENGINEERS
In the testing program described in this report, supervised by John A.
Blume & Associates, Engineers, fatigue tests related to first crack
load were carried out. These tests and results are discussed in
Chapter IV.
7. Impact Resistance
The behavior of ferro-cement under impact loading is typified by high
ductility and energy absorption in the plastic range, good localization
of damage, and generally easy repair of damaged areas. Experimental
testing has been done to compare impact resistance of ferro-cement with
that of reinforced concrete, fiber reinforced plastic, and marine
plywood. Material parameters affecting impact resistance have been
found to be thickness of cross section, volume and surface area of
reinforcement, and strength of reinforcement.
Impact studies in the Soviet Union reported in 1968 by V. F. Bezukladovl l
compared ferro-cement to reinforced concrete. Tests involved dropping
a 10-inch sphere weighing 55 pounds on 20-by 36-inch panels of 1-inch
thick ferro-cement and 2-inch thick reinforced concrete. Results
indicated that the ferro-cement performed slightly better.
Tests conducted in 1970 at MIT by S. P. Shah and W. H. Key, Jr.2 0
evaluated the effect of specific surface of reinforcement and strength
and ductility of reinforcement on impact resistance. Tests involved
striking 9-inch by 9-inch by 1/2-inch thick panels with a ballistic
pendulum and measuring absorbed energy and damage to the panel. Damage
was evaluated by measuring the leakage flow rate of water under a
constant head through the impacted region. Steel reinforcement percent-
age was held constant while ductility and specific surface were varied.
Typical results are shown in Figure 13. Both high specific surface of
reinforcement and high tensile strength, low ductility steel resulted
in lower leakage rate and thus, by definition lower impact damage for
a constant value of absorbed impact energy.
39- 20 -
JOHN A. BLUME & ASSOCIATES, ENGINEERS
Rio .as
I 3 4
TENJILE YIEL.O LOAM: OF REI/FOIeCEMENT, /t6 ,t/O(VOLcUME OF RIE//NIOIRCEMe'r'T 'J CONJSTANr)
FIGUR. 13 - EF/fECT OF JEeCI/FIC JURFACEANOv '7/NJ7" ,E STRNE-G7W OFm e"A/,FOZ CEM ,NA7-oCN /MPACT #OAMAGe '( FRoM Rei. go20)
JOHN A. BLUME & ASSOCIATES, ENGINEERS
-. 40I
,-.1L
O'u
20U 1,
Ou c
Jr a 2.5
i _
I ,
Comparison with other materials and a method for improving impact re-
sistance were studied in 1971 by K. A. Christensen and R. B. Williamson
at the University of California, Berkeley.2 1 Panels of ferro-cement,
marine plywood, and fiber reinforced plastic (FRP) ranging in thickness
from 1/2- to 1-1/2-inches were tested under impacts from a 25-pound
hemispherical weight dropped from heights up to 18 feet. Damage was
measured by the leakage flow rate of water through the damaged area.
The primary focus of the tests was on establishing a single-blow
"critical impact" for each type of panel that resulted in a leakage
rate of 6 gallons per hour under a 2-foot head, which was defined as
a "critical condition" of impact damage. Test results comparing ferro-
cement to marine plywood of equal thicknesses showed that ferro-
cement impact resistance varies from 50 percent less than the plywood
for thin (1/2-inch) sections to approximately equal to the plywood for
thicker (1-1/2-inch) sections. Results showed the FRP to be far
superior, the 1/2-inch FRP section having 100 percent more impact
resistance than the 1-1/2-inch ferro-cement section.
The test results by Christensen and Williamson showed that ferro-
cement, similar to fiber reinforced plastic, exhibits good plastic
absorption of impact energy after partial failure. A typical impact
specimen was described as follows:
The top surface of the impacted ferro-cement specimensgave little evidence that the panel was in a criticalcondition. A smooth crater in the mortar without exposureof the mesh did not indicate that passage of water mightresult. Damage to the impacted surface was confined tocrushing directly under the impact device. On the sideopposite the impacted surface crack propagation wasarrested within a small area, approximately three timesthe diameter of the impacting device. This was accom-panied by approximately 1/2 inch bulging with virtuallyno loss, spalling, or breaking away of the mortar, eventhough there was cracking and loosening of the mortarin the mesh. The mesh restrained and retained thebroken mortar.2 1
41- 21 -
JOHN A. BLUME &8 ASSOCIATES. ENGINEERS
A sandwich panel designed to improve impact resistance was developed
as part of their ferro-cement impact studies. This new section consists
of ferro-cement overlaid with fiber reinforced plastic. A 3/4-inch
sandwich panel consists typically of 1/2-inch of ferro-cement overlain
with 1/4-inch of FRP. Test results showed impact resistance increased
eight to ten times compared to equal thicknesses of normal ferro-
cement. The materials and fabrication of these sandwich panels would
be more costly than plain ferro-cement, yet this material provides a
good alternative in critical impact areas of ferro-cement structures.
Christensen and Williamson note that no theoretical consideration of
ferro-cement impact resistance exists and none was attempted in their
report. Their work was intented to provide a guide to ferro-cement
impact strength by comparison with the more widely known and documented
materials, marine plywood and fiber reinforced plastic. Relative to
steel, the impact strength of ferro-cement in resisting complete failure
(i.e., punch-through) has been estimated by R. B. Williamson as that of
a steel plate containing the same volume of steel as the total volume
of the ferro-cement reinforcement.8 Although ferro-cement impact
strength is low relative to steel and fiber reinforced plastic, good
ductility of ferro-cement in absorbing impact energy after partial
damage, as well as ease of repair of damaged areas, appear to make
ferro-cement feasible for use in many structures where impact loads
are expected.
Because of the limited impact strength of ferro-cement, tests should
be developed to approximate actual anticipated impact loads, such as
vehicle collision or propeller blade impact. Various ferro-cement
sections should be compared and the need for strengthening such as
FRP overlay or steel plate backing evaluated.
42- 22 -
JOHN A. BLUME & ASSOCIATES, ENGINEERS
B. MATERIAL CONSTITUENTS
The basic ferro-cement materials -- mortar and reinforcement -- are
discussed in the following sections. The types and proportions of
these materials and their effect on various ferro-cement properties
are described.
1. Mortar
The nature of ferro-cement mortar and the influence of various mix
design parameters on ferro-cement mortar are important considerations.
As in reinforced concrete construction, mix design in ferro-cement
affects a number of material characteristics including compressive
strength, drying shrinkage, density, and durability. The Portland
cement mortars used in most reported ferro-cement construction projects
and laboratory tests can be generally characterized as cement-rich with
relatively low water content and high strength.
The types of cement used in various ferro-cement projects vary widely
depending on the need for such qualities as low shrinkage, high early
strength, or high resistance to corrosion. The use of sulfate-
resistant cement is desirable for increased corrosion resistance as
is low alkali cement for reducing the detrimental effects of alkali-
aggregate reaction. Aggregates are generally fine sand, although some
recent research and construction projects have used manufactured
lightweight aggregates to give better uniformity of aggregate and
reduced structural weight.2 2'4
The proportion of cement in ferro-cement mortars is generally high,
providing good workability and high density. Typical values for sand-
to-cement ratio are from 1.0 to 2.0, although high cement content can
lead to problems in drying shrinkage. Values of water-to-cement ratio
from 0.33 to 0.60 have been used, although a low value is desirable
_43- 23 -
JOHN A. BLUME & ASSOCIATES, ENGINEERS
for high strength and low permeability. Zero-slump mortar has been
used in conjunction with molds and vibration during placement.
Additives in ferro-cement mortar have been used primarily to improve
such properties as workability and resistance to chemical attack, and
to reduce drying shrinkage. The use of such additives is well-documented
in concrete literature. The most commonly used have been pozzolan or
fly ash to improve workability.
A significant contribution to ferro-cement technology has been the use
of a chromium trixoide additive by K. A. Christensen and R. B.
Williamson, Department of Civil Engineering, University of California,
Berkeley.2 3 Where bare steel reinforcing bars and galvanized wire
mesh are used together, a galvanic cell action is set up between the
steel and zinc mesh coating with the wet mortar acting as the electrolyte,
causing hydrogen gas bubbles to be released. These bubbles seriously
disrupt the surface quality and the bond between the mortar and rein-
forcing bars. The addition of chromium trioxide to the mortar mix
largely inhibits this electrochemical reaction.
A number of ferro-cement characteristics including strength, drying
shrinkage, durability, and economy are greatly affected by mix design.
Until such time as comprehensive design procedures are developed,
detailed study of mortar mix design is recommended for major structural
applications. Because of the typically thin cross section and rein-
forcement coverage typical of ferro-cement, durability is especially
dependent on good mix design. The subject of corrosion resistance is
discussed further in a later section.
2. Reinforcement
The preceding sections covering the state-of-the-art in ferro-cement
strength properties included considerations of reinforcement. For
44JOHN A. BLUME 8& ASSOCIATES. ENGINEERS
clarity the various parameters related to ferro-cement reinforcement
are summarized and discussed in this section.
Two commonly used reinforcement configurations are shown in Figure 1.
Section la of Figure 1 shows the use of reinforcing bars in conjunction
with wire mesh. The bars are at the center of the ferro-cement cross
section and the wire mesh is laid over each side of the network of bars.
Reinforcing bars are typically 1/4-inch bars on 2-inch centers while
wire mesh sizes have varied greatly in recent test programs and
construction projects. Typical mesh sizes have 1/2-inch or 1/4-inch
wire spacing and utilize 16-to 24-gage wire.
Section lb of Figure 1 shows a ferro-cement section using only wire
mesh reinforcement. Typical mesh sizes are similar to those mentioned
above. Reinforcement coverage for both sections is typically 1/16-to
1/8-inch. To maximize reinforcement volume and specific surface, the
maximum number of mesh layers generally are used consistent with main-
taining the required coverage within a given thickness.
Reinforcement parameters related to the strength properties and other
performance characteristics of ferro-cement sections include the
following:
* Volume and specific surface of reinforcement
* Size of reinforcement (diameter, spacing)
* Strength and ductility of reinforcement
* Type of reinforcement (reinforcing bars, welded wire mesh,
The most significant reinforcement parameters are volume and specific
surface. Increased volume percentage of reinforcement is related
closely to increased tensile stress at first crack (Figure 4) and
JOHN A. BLUME & ASSOCIATES, ENGINEERS
increased tensile stress at ultimate load (Figure 7). The results of
bending tests have related increased volume of steel to increased
ultimate flexural capacity.1 4 Volume percentage of reinforcement
directly affects the elastic modulus of ferro-cement, increased
volume resulting in increased modulus (Figure 11). It appears obvious
that other ferro-cement properties such as fatigue resistance are re-
lated to volume of reinforcement, although this has not been experi-
mentally verified as yet.
The specific surface of reinforcement is the most sensitive material
parameter in relation to strain and cracking behavior. It should be
noted that a high value of specific surface is the principal character-
istic that separates ferro-cement from reinforced concrete. The use
of reinforcing bars in ferro-cement improves some material strength
properties through increases in reinforcement volume, but the only way
of obtaining significantly high specific surface is through the use of
wire mesh reinforcement, or similar materials such as expanded metal
lath, which provide good reinforcement dispersal.
Increased ferro-cement stress at first cracking for both flexure and
tension is very closely related to increased specific surface (Figures
5 and 9). Impact resistance is similarly related to specific surface
(Figure 13). Several ferro-cement research reports have stated that a
high value of specific surface is one of the primary factors that
distinguish it from reinforced concrete. I. R. Walkus and T. G.
Kowalski6 state that a specific surface of reinforcement greater than
about 2.5-inch2 /inch3 delineates ferro-cement from ordinary reinforced
concrete.
The size and spacing of reinforcement are the primary variables in
establishing the value of the specific surface. In general, smaller
diameter wires spaced closer together result in higher specific
surface of reinforcement. The ease of construction is also affected
by size and spacing of reinforcement. The use of more widely spaced
wire mesh allows easier and better mortar penetration.
- 26 -
46 JOHN A. BLUME & ASSOCIATES, ENGINEERS
Because the ultimate strength of ferro-cement in tension and flexure
is directly related to reinforcement capacity, the strength of rein-
forcement is similarly related to these material properties. A higher
strength (less ductile) steel results in higher load and smaller cracks
at ultimate load. Higher impact resistance (Figure 13) is also related
to higher strength steel.
Several kinds of reinforcement have been successfully used in ferro-
cement construction. The use of steel reinforcing bars along with wire
mesh gives increased flexural stiffness and increased flexural yield.
The inclusion of reinforcing bars also has been observed to improve
shear strength. The most commonly used wire reinforcements are woven
wire mesh and welded wire mesh. Expanded metal lath and chicken wire
have also been used. Comparative studies of various reinforcements
have shown certain weaknesses in some types. Woven wire mesh and ex-
panded metal lath reinforcement result in variation in the ferro-cement
tensile properties in the two orthogonal directions. For woven wire
mesh with large weave angle and also for chicken wire reinforcement,
spalling of the mortar matrix has been observed at the ultimate load
of the reinforcement. It should be noted, however, that the above
materials have been used successfully in marine applications where
service loads are well below ultimate value. The use of expanded
metal lath can, in fact, result in very high values of specific surface.
Ungalvanized wire mesh has been found to provide greater ultimate
strength because the process of galvanizing anneals and weakens wire.
Also, ungalvanized woven wire mesh has been found to be desirable
when tight curvatures or double curvatures are encountered because
there is no bonding between orthogonal wires and it is therefore
easier to form. However, except when protective surface coatings are
used, galvanized reinforcement is generally desirable for increased
corrosion protection because of the relatively thin mortar coverage
over the reinforcement in most ferro-cement sections.
-27 4 7
JOHN A. BLUME & ASSOCIATES, ENGINEERS
Optimization of the reinforcement for a specific structural application
is a complicated problem and is related to other factors including
magnitude and type of loads, fabrication methods, service environment,
required service life, strength and stiffness required, and economic
considerations.
Tancreto and Haynes have concluded, based on their studies of the flexural
behavior of ferro-cement with various sizes of woven wire mesh reinforce-
ment, that the best overall compromise for strength, workability, and
cost was a 1/4-inch by 1/4-inch mesh of 23-gage wire. 1 4 However, com-
prehensive data for all types of loading and reinforcement are not yet
available. Therefore, a detailed reinforcement optimization study in
conjunction with a mortar mix design study is recommended for any major
ferro-cement structures.
C. PERFORMANCE CHARACTERISTICS
The following sections contain discussions of various ferro-cement
characteristics which are generally independent of strength properties.
These include surface finish, durability, acoustical and fire resistance
characteristics, repairability, dimensional stability, and maintenance.
1. Surface Characteristics
The surface characteristics obtainable in ferro-cement construction are
similar to those for reinforced concrete or precast concrete construction.
Surface control can be achieved for just about any level of accuracy
through special fabrication methods, such as precision molds. Almost
any degree of surface smoothness can be achieved by treatment of the
molds or by finishing techniques. In addition, mortar additives and
surface coatings can be used to improve smoothness and durability.
Using present mechanical or hand finishing techniques for concrete
floor slabs, surface textures ranging from very rough or skid resistant
- 2848JOHN A. BLUME & ASSOCIATES. ENGINEERS
to extremely smooth can be obtained. By using smooth molds and vibration
during placing, surface texture on the contact surface can be made hard
and smooth.
Surface durability is largely dependent on the quality of the mortar.
Reports by the Portland Cement Association2 4 and the American Concrete
Institute2 5 on the wear resistance of Portland cement concrete show that
compressive strength is the most important single factor related to wear.
These studies show a relationship between increased concrete compressive
strength and increased abrasion wear resistance. A similar relationship
was observed for mortar, although abrasion wear resistance of mortar was
found to be less than that of concrete with large aggregate. Wear resis-
tance of concrete and mortar is also enhanced as cement content is
increased and water-to-cement ratio is decreased. Both of these factors
help to make the mortar matrix more dense and favor the wear resistance
of ferro-cement mortars that are characteristically cement-rich and low
in water content.
Surface durability of concrete and also of ferro-cement is dependent
on finishing and curing. Excess surface moisture and rapid loss of
surface moisture must be avoided. These conditions can be met within
the scope of normal concrete technology. Epoxy coatings and marine
paints have been used on ferro-cement boat hulls for added corrosion
protection and surface durability. Coatings of this type could add
significantly to costs, yet their use may be required in certain
applications. The need for surface coatings is further discussed in
the section on corrosion resistance.
Studies conducted in 1970 at MIT by S. A. Frondistou-Yannas and
S. P. Shah2 6 show that the addition of polymer latex additives to
ferro-cement mortar result in higher extensibility and greater toughness.
These results imply that wear resistance could be enhanced by the use of
this type of additive in ferro-cement mortars. However, Frondistou-
JOHN A. BLUME & ASSOCIATES, ENGINEERS
Yannas and Shah show that the addition of polymer latex additives to
ferro-cement mortar result in higher extensibility and greater
toughness. These results imply that wear resistance could be enhanced
by the use of this type of additive in ferro-cement mortars. However,
Frondistou-Yannas and Shah observed that polymer latex additives had the
detrimental effect of causing more shrinkage; thus this area warrants
further study.
Although no data exist on the effects of high-velocity winds, such as
occur in wind tunnel structures, Portland Cement Association studies2 4
of hydraulic abrasion and cavitation show that good quality concrete is
not affected by a steady, tangential, high-velocity flow of water. The
long-term behavior, however, of plain and polymer latex-added or epoxy
coated ferro-cement mortars in wind tunnel environments should be
studied. A preliminary evaluation of the effects of high-velocity air
flow on ferro-cement was conducted in the test program described in this
report. Findings are discussed in Chapter IV.
2. Corrosion Resistance
Factors affecting ferro-cement durability and methods for improving it
are discussed in this section. Factors related to corrosion and
corrosion resistance of ferro-cement can be obtained from sources
within the concrete industry. Information about concrete durability,
concrete protection, and the use of admixtures relative to corrosion
resistance is included in the reports of ACI Committee 201,25 ACI
Committee 515,27 and ACI Committee 212,28 respectively.
Deterioration of reinforced concrete (and ferro-cement) in corrosive
environments generally results from chemical attack on the concrete and
corrosion of the steel reinforcement. The extent of chemical attackon
concrete is mainly related to its permeability, its alkalinity, and
the tendency of hydrated cement compounds to undergo undesirable
chemical reactions such as sulfate reactions. Penetration of fluids
JOHN A. BLUME & ASSOCIATES, ENGINEERS
into concrete can cause adverse chemical reactions with the cement,
aggregate or steel. Corrosion of steel reinforcement in concrete
results primarily from penetration of corrosive agents through cracks
in the concrete covering, and deterioration of the concrete cover. The
degree of protection given reinforcement by the concrete cover is
dependent on the concrete quality and depth of cover.
Susceptibility of ferro-cement to deterioration from corrosive elements
is primarily related to reinforcement cover, which is typically quite
thin (about 1/8 inch) compared to ordinary reinforced concrete, and
therefore provides less protection to the reinforcement. This problem
has generally been treated by the following measures: (a) Increasing
impermeability of'mortar by low water-to-cement ratio, and careful
attention to proper proportioning, grading, mixing, placing, and curing;
(b) Application of impermeable coatings; (c) Added protection to the
reinforcement such as galvanizing; and (d) Proper design to eliminate
or minimize the extent and width of mortar cracks. A qualitative re-
lationship between crack width and corrosion resistance, established
by Walkus and Kowalski,6 is shown in Figure 3. In general, dealing
with the susceptibility to corrosive attack resulting from the thin
ferro-cement section and reinforcement cover requires high quality
materials for the mortar and a high degree of control in mixing and
placing.
The need for protective coatings is based on the specific structural
application. The following two general conditions exist relative to
corrosion resistance for reinforced concrete and ferro-cement:2 5
1. Those in which proper attention to the concrete itselfwill provide immunity or an acceptably low rate ofdeterioration, and
2. Those in which it is necessary to prevent contact betweenthe corrosive chemical and -the concrete by means of aprotective coating.
si- 31 -
JOHN A. BLUME & ASSOCIATES. ENGINEERS
Concrete and mortar specifications that assure resistivity to many
corrosive elements are well within the state-of-the-art in concrete
technology. For important or unusual uses of ferro-cement, however,
especially in view of the thin reinforcement coverage, new specifications
should be developed that include tests to determine adequacy of corrosion
resistance. In case of inadequacy, protective coatings may be desirable.
Experience with ferro-cement in marine construction has shown that
application of impervious coatings is required in most cases.
Protective coatings that have been used in marine ferro-cement con-
struction are primarily resin-based coatings such as polyester,
urethane, and epoxy. Desirable qualities for coatings include low cost,
ease of application, impermeability, chemical resistivity, extensibility,
and toughness. Careful preparation is necessary for applications of
these coatings. A significant problem, especially in amateur-built
ferro-cement boats, has been debonding of protective coatings.
Additives have been used in ferro-cement mortar to obtain greater
density and hence greater corrosion resistance. One commercial boat
building company employing ferro-cement is currently using an acrylic
latex additive with apparent success.2 9 The outside (or gel) coat of
mortar contains the acrylic latex additive while the remainder of the
cross section uses regular mortar. Experiments conducted at MIT2 6
showed that polymer latex additives produce mortar of increased
toughness and extensibility, although increased drying shrinkage
resulted. Based on these findings and the above mentioned use in
boats, the use of a polymer latex added gel-coat appears very promising
for many kinds of ferro-cement construction. The expense of applying a
protective coating could be eliminated as well as the possibility of its
debonding. This area should be carefully considered in future ferro-
cement studies.
Studies of the effect of freeze-thaw cycles on ferro-cement samples
conducted by A. M. Kelly and T. W. Mouat3 0 are shown in Figure 14.
- ' JOHN A. BLUME & ASSOCIATES, ENGINEERS
53
.
0
.
.
50 /00 /50 200 250 300
MUM 1 ER OF FRCEZE -7WAW CYCLES
FIG UR E /4 - FAREEZE - T7A W C CL REF. 80)
JOHN A. BLUME & ASSOCIATES; ENGINEERS
SO
40
30
k
Lb
w
u
K0
U
(3
ol
qI,
0
1 t\U
'10
TE JTJ
CFROM
These tests were conducted in accordance with ASTM C 291-67 (Method of
Test for Resistance of Concrete Specimens to Rapid Freezing in Air and
Thawing in Water) on 3/4-inch ferro-cement specimens of three types
containing one, six, and twelve layers of mesh. The beneficial effect
of adequate reinforcement is evident from these tests. It should be
noted that the test procedure for ASTM C 291 is generally more severe
than freeze-thaw cycles in actual structures.
Ferro-cement corrosion resistance is an important area for future study.
Experience with ferro-cement marine craft shows that resistance to a
corrosive sea water environment can be obtained. However, various forms
of protective coatings are required for this. Because this corrosion
protection adds significantly to costs, the economic feasibility of some
proposed ferro-cement structures can be affected. Comparative studies
of corrosion resistance for various types of ferro-cement, in conjunction
with mortar mix and reinforcement studies are recommended in relation to
specific structural applications.
3. Vibration and Acoustical Characteristics
No test results on vibration and damping of ferro-cement have been
reported at present. It is expected that vibration characteristics
would be approximated by the law of mixture of composite materials
similar to Shah's results for modulus of elasticity. Vibration and
damping have been studied in the test program described in this report
and are discussed in Chapter IV.
Although no tests on the acoustical properties of ferro-cement have
been reported, approximations can be made based on work with concrete.
Figure 15 shows the sound-transmission loss through dense concrete and
steel based on an approximate design method described by I. L. Ver and
C. I. Holmer.3 1 It can be seen that 5/8-inch and 1-inch thick dense
concrete exhibit greater attenuation than 1/8-inch and 3/8-inch steel
plate over some portions of the sound frequency range. These results
JOHN A. BLUME & ASSOCIATES. ENGINEERS
55
3^ 60X is COCRC-7 C
6'/ SNJ~ CONC~e
0
$0 J
/U,
Zo
6 to/ otI io lo
JouNO FREQUEN'CY , HLc
PI=URE IS- £OUN0I7-ToANJMIj/l-ON LOSJ l=OS
/o0
JOHN A. BLUME & ASSOCIATES, ENGINEERS
:-OENJ -CONCRETET (PEtRO' CM/,eA7).. - Y...reL (Ci=Om.o e, St )
56
give an approximate comparison between steel and ferro-cement with
regular weight aggregate. However, conditions in an actual wind tunnel
could vary. The acoustical properties of lightweight aggregate ferro-
cement and those of ferro-cement with tensile or flexural cracks would
probably vary from that of dense concrete. Comparative attenuation
tests of various ferro-cement configurations in actual wind tunnel
environments would be especially significant.
4. Fire Resistance
Ferro-cement has high fire resistance because the cement mortar is not com-
bustible and provides insulation for the steel reinforcement. Tests con-
ducted by Windboats, Ltd., of England in 1963 indicate that "test panels
have withstood 17000 C for 1-1/2 hours with no effect on the material."1 8
However, heat does have a certain deleterious effect on cementitious materials
and in addition, the 1/16-to 1/8-inch reinforcement coverage typical of
ferro-cement is well below the minimum coverage required in reinforced con-
crete structures. These factors should be- considered in experimental work
to establish a fire rating for various ferro-cement configurations.
5. Repairability
The lower impact resistance of ferro-cement relative to fiber reinforced
plastic and most metals is offset to a large extent by its ease of
repair. Literature on ferro-cement marine craft contains a number of
accounts of repair of severe impact damage within a few hours at nominal
cost either in port or under way. Procedures for ferro-cement repair
are similar to those for concrete and can be obtained from sources
within the concrete industry such as the American Concrete Institute
Manual of Concrete Practice.3 2 Typical repair materials are Portland
cement mortar, epoxy grout, or commercial patching mixtures. Repair
procedures involve removing damaged or spalled mortar, straightening or
replacing deformed reinforcement, treating broken surfaces with an
etching or bonding agent, and applying new mortar or patching material.
- 34 -
JOHN A. BLUME & ASSOCIATES, ENGINEERS
57
Repair of ferro-cement adversely affected by corrosion or fire would
be done in a similar manner.
Tests conducted in 1970 by A. M. Greenius and W. N. English3 3 for the
Canadian Department of the Environment evaluated the strength of various
ferro-cement repair materials. One-inch thick ferro-cement panels were
tested to failure under flexural and impact loadings, repaired with one
of several patching materials and then retested to failure. Patching
materials used were Portland cement mortar consisting of one part cement,
two parts sand and 0.4 parts water by weight; a commercial epoxy marine
patching compound; and a commercial epoxy floor patching material.
Test samples repaired with Portland cement mortar were retested after
21 days curing time. Samples that had suffered reinforcement damage
during original failure retested at 50 percent to 70 percent of their
original strength. Other samples retested at about 80 percent of their
original strength.
Test samples repaired with epoxy materials were retested after 7 days
and regained virtually all their original strength. The repair with
epoxy was affected by forcing the patching material into the cracks
produced by initial failure, rather than chipping away damaged material
as in the case of the repairs using Portland cement mortar. On re-
testing, failure occurred away from the epoxy repaired cracks.
6. Dimensional Stability
The primary factors influencing structural dimensions and tolerances
are drying shrinkage and creep. Drying shrinkage can be minimized by
proper mix design and is a relatively short-term phenomenon. When a
structure is cast in segments it can generally be compensated for by
casting the segments slightly oversize. In the case of cast-in-place
ferro-cement used in conjunction with precast concrete ribs, special
- 35 -
JOHN A. BLUME & ASSOCIATES. ENGINEERS
58
care should be taken to minimize shrinkage which results in internal
stress build-up.
Ferro-cement creep has been studied in testing work reported by the
Naval Ships Research and Development Center.l9 The results are shown
in Figure 16 and are for 1/2-inch ferro-cement specimens containing
6.2 percent galvanized wire mesh and loaded in flexure. The report
of this work showed that from the nature of the log-log plot of the
data that the ferro-cement samples tested have "the creep characteris-
tics of a metal and not those of a composite."1 9 The data shown in
Figure 16 are for a limited time duration. For structures with large
dead loads or operational loads, creep data over a more extended time
period are needed.
As in concrete construction, the use of prestressing in ferro-cement
or in concrete ribs cast integrally with ferro-cement should be care-
fully studied from the point of view of dimensional changes caused by
creep. Creep of prestressed members generally occurs over an extended
period and can cause problems in the areas of joints and connections.
7. Maintenance
Maintenance of ferro-cement is discussed in this section, with reference
to wind tunnel structures. Published data on long-term maintenance
problems encountered in non-marine ferro-cement structures are not
available. Long-term maintenance problems in ferro-cement marine craft
are not well defined because almost all ferro-cement boats currently
operating have been built since the mid-1960's.
The most important maintenance problem in ferro-cement boats is main-
taining adequate protection against chemical (especially sulfate)
attack. The prolonged exposure of ferro-cement boat hulls to sea
water has generally required the use of protective coatings. Most
failures of the coating material or of the hull itself have resulted
- 36 -
JOHN A. BLUME & ASSOCIATES, ENGINEERS
.001
- 0001
.000
.00001
rlME, HOUfr
CF/UeE /J - FeeOo- CIME.N7 CIEEP T)ESTCFROM REF. I 9)
JOHN A. BLUME & ASSOCIATES, ENGINEERS
59
09
Si
I /0 t00 /000
<~~~~~~~~.
I
from poor workmanship in obtaining good mortar penetration into the
reinforcement and from poor selection of materials for reinforcement
or coatings. Inadequate mortar penetration or high permeability have
led to excessive deterioration from freeze-thaw cycles or sulfate
attack.
Reinforcement configurations that allow excessive cracking have re-
sulted in rapid sea water penetration and resulting deterioration. Some
epoxy coating materials have been found to be inadequate. Coefficients
of expansion for temperature and moisture change, which are much greater
than that of the ferro-cement mortar, have led to shear failure in the
mortar just under the coating. Debonding of the coating has occurred
in many cases within 1-to 3-years.
Instances of failure and debonding of surface coatings in ferro-cement
boats have been troublesome to repair, but experience has shown that
such coating failures do not damage well-made ferro-cement hulls.2 9
As previously discussed, in the absence of protective coatings the most
important factor in obtaining ferro-cement resistance to chemical
attack is high-quality mortar with low permeability.
Ferro-cement structures in atmospheric environments rather than sea
water should be subjected to a much lower rate of chemical attack.
Many structures -- with high-quality mortar, adequate reinforcement
cover, and good design minimizing deflection and cracking -- can be
used without coatings and be expected to require no maintenance.
Where moderate chemical attack is expected, such as from air pollution
or engine exhaust residues, protective coatings may be required. Based
on recommendations from persons currently engaged in ferro-cement
research and construction,8 ' 2 9 a good solution appears to be the use
of a gel-coat of mortar with a polymer latex additive. This was pre-
viously discussed in the section pertaining to corrosion resistance.
- 3760JOHN A. BLUME & ASSOCIATES. ENGINEERS
The use of this gel-coat provides increased density and chemical resis-
tance in the outer layer of the ferro-cement mortar and decreases the
problems of debonding and mortar cracking. Although there are no long-
term test results on the use of this gel-coat, based on the present
state-of-the-art it appears to be the best way of providing a minimum
maintenance ferro-cement structure for wind tunnel construction. Current
and future research on protective coatings may produce improvements in the
use of gel-coats and also other kinds of protective coatings.
Future ferro-cement test programs should include studies to verify the
adequacy of the polymer latex additive gel-coat and studies to compare
the basic properties and the effectiveness of various protective coatings.
For wind tunnel structures subject to seve're vibration loads, the current
fatigue studies should be extended to determine the long-term effect on
surface properties and maintenance problems.
D. CURRENT CONSTRUCTION METHODS
The following section constitutes the second part of the state-of-the-
art study of ferro-cement where current construction methods were
studied which can be applied to wind tunnel structures. The most common
methods have been developed primarily for marine construction. These
methods are adaptable for fabrication of other kinds of structures,
including large wind tunnel structures. Of the commonly used construction
methods, the two principal ones differ basically in the means by which the
shape of the finished structure is formed and controlled. In the first
method, structure shape is defined by a network of steel reinforcing
bars which is overlain with wire mesh (Figure 1). The surface is
controlled by hand placing and troweling of the mortar. The second
method utilizes a mold to achieve the required finished shape and surface.
The use of precisely constructed, reusable molds appears to be especially
compatible with the structural and economic requirements of wind tunnel
construction.
- 38
JOHN A. BLUME & ASSOCIATES, ENGINEERS
Ferro-cement construction using frames or networks of reinforcing bars
for shape control is more labor-intensive and achieving close dimensional
tolerance is more difficult. The cross section shown in Section la of
Figure 1 is typical of one-unit, amateur-built private yacht construction
where the cost of a mold for one unit is excessive and where labor time
is not accounted. A large amount of labor is required for fabricating
the reinforcing bar and wire mesh reinforcement network, placing the
mortar to insure proper penetration, and trowel-finishing the mortar to
the required dimensional precision.
The problem of dimensional tolerance can be greatly reduced by the use
of molds. The use of molds also reduces labor in placing and finishing
the mortar and generally requires less skilled workmen. Molds are
currently being used in the construction of a large ferro-cement barge
in Vancouver, British Columbia, and of large sailboat and motorboat
hulls in West Sacramento, California. These two projects warrant
further discussion because they involve several of the latest innovations
in large-scale, commercial ferro-cement construction.
The construction of a 180-foot prototype ferro-cement cargo barge by
FERROCON Industries, Ltd., Vancouver, British Columbia,4 has been
partially sponsored by the government of Canada for the purpose of
developing new Canadian technology. A large amount of research was
conducted in connection with this project and the test data has been
retained as proprietary information by FERROCON Industries. Discussions
with David J. Seymour, Naval Architect,3 4 who had complete design
responsibility for the FERROCON barge project, indicate that some of
the most advanced state-of-the-art procedures in ferro-cement design
and construction were utilized. Precast, post-tensioned concrete ribs
and frames were used in conjunction with cast-in-place ferro-cement
shells for the hull and deck. The vessel was designed to comply with
the standards of the American Bureau of Shipping for ocean-going
barges. As an example of the strict specifications, a minimum modulus
of rupture of 6000 psi was required for the hull and deck. Zero-slump
JOHN A. BLUME & ASSOCIATES, ENGINEERS
mortar and extensive vibration during placement of mortar were used to
help meet this strength requirement. It is expected that most design
requirements for civil engineering uses of ferro-cement will be below
those encountered in this project.
The 55-foot motor and sailboat hulls being produced by Fibersteel
Corporation, West Sacramento,2 9 utilize a concrete cavity mold for the
hull and deck and also a unique process for casting ferro-cement. By
this process, a wet-mix spray gun is used to cover the surface of the
cavity mold with a 1/16-inch gel-coat of ferro-cement mortar containing
an acrylic latex emulsion additive. After this first coat has an
initial set, a first layer of reinforcement is laid onto the mortar and
a coat of ordinary mortar is sprayed on. A second layer of reinforcement
is then placed before the mortar sets and the sequence is continued until
the required thickness is reached. Some advantages of this method are
good surface and dimensional control through the use of molds, excellent
penetration of mortar into reinforcement mesh, hence good bond without
the need for vibration and good adaptability for production line
operations. Compared to the methods developed by FERROCON Industries
this method uses a wetter, thus somewhat weaker, mortar mix and gives
less accurate placement of reinforcement. However, for uses where load
levels are relatively low and ease of construction becomes of primary
importance, it appears that a process similar to this layering method
would result in cost advantages.
A third significant ferro-cement project is currently being carried out
by the Naval Ship Research and Development Center.5 This work involves
research and prototype construction of 24-foot ferro-cement motor
launches. Steel molds and hand application of mortar were used for
construction of three prototypes. Minimum weight was important for
these craft and hulls as thin as 3/8 inch were used. At least one
prototype was successfully field tested for a year in sea conditions
exceeding design requirements.
46 3JOHN A. BLUME & ASSOCIATES, ENGINEERS
in the survey reported by Walkus and Kowalski,6 which is related primarily
to current ferro-cement work in Eastern Europe, some interesting fabrica-
tion techniques are discussed. These techniques have been developed
primarily for prefabricated curved or folded-plate roof elements and are
referred to as "vibro-pressing" and "vibro-bending." They are described
as follows:
Basically, "vibro-pressing" consists of a concrete dispensermoving on rails and following the profile of the element,the dispenser vibrating and pressing the mortar into the meshfixed below it. "Vibro-bending" usually means that a ferro-cement sheet is cast flat on a steel mold which then has itssides raised to form finally a V-shape. The bending isaccomplished through a small angle and to a relatively largeradius so as to create the least disturbance for the partsalready cast. Another variation of this technique consistsof "winding" a ferro-cement sheet on a circular former so asto produce a trough element. The process may also beaccompanied by mold vibration.6
Material and fabrication costs for the large ferro-cement marine projects
currently under way in the United States and Canada are difficult to
evaluate because most are in preliminary or prototype stages where costs
are not representative of final construction. The feasibility of ferro-
cement for civil engineering structures is expected to depend heavily
on the economy of fabrication. The application of ferro-cement to wind
tunnel construction is discussed in Chapter V. In Chapter VI various
cost factors related to this kind of construction are reviewed.
64- 41 -
JOHN A. BLUME & ASSOCIATES. ENGINEERS
IV. FERRO-CEMENT TESTING PROGRAM
The review of published and unpublished data discussed in the preced-
ing chapter indicated that in a number of areas insufficient informa-
tion was available. A limited test program was therefore developed
and conducted to obtain preliminary data about those parameters that
could be important in wind tunnel construction. The ferro-cement
characteristics tested are related to the high speed air flow and
vibratory loading present under operational conditions in most wind
tunnels and include: strength and cracking behavior under fatigue
loading, vibration and damping properties, and resistance of ferro-
cement surfaces to abrasion from high velocity air flow.
Inadequate performance in any of the above material characteristics
could seriously limit the feasibility of ferro-cement for wind tunnel
construction. Therefore, to improve the reliability of the conclu-
sions reached in this report, tests were conducted to obtain prelim-
inary data in these areas. The tests were designed to provide pre-
liminary values of the material properties from which estimates as
to the adequacy or inadequacy of the material could be made. In ad-
dition, some static tests in compression, tension, and flexure were
conducted to provide a correlation between the properties of the test
samples studied in this program and those studied by other investiga-
tors as described in the preceding chapter. The materials and pro-
cedures utilized in testing are described in Section A and the results
are presented and discussed in Section B of this chapter.
A. DESCRIPTION OF TESTS
Criteria for the tests were developed by John A. Blume & Associates,
Engineers. Test specimens were furnished to Dr. William J. Venuti,
Professor of Civil Engineering, California State University, San Jose,
JOHN A. BLUME & ASSOCIATES, ENGINEERS
who conducted the tests. The tests were performed in the Advanced
Structures Laboratory at the University. Appendix C contains a de-
tailed report of the test equipment and experimental methods.
1. Test Samples
Two large ferro-cement panels were fabricated by Fibersteel
Corporation, West Sacramento, using the layering process described
in the preceding chapter. The panels were 3 feet by 5 feet by
1/2-inch thick and were made in accordance with the specifications
listed in Table 1. The panels were cast on 3/4-inch concrete
form plywood and after curing were cut into test samples. Fig-
ures 17a and 17b show the mortar spraying process and a layer of
wire mesh placed into the wet mortar. The casting procedures and
materials are the same as those used by Fibersteel in fabricating
ferro-cement boat hulls, although they normally use expanded metal
instead of wire mesh. Because of the fabrication procedures used
test results should be representative of the properties obtainable
in actual field-fabricated ferro-cement.
The panels were cured under a combination of steam and ambient
temperature conditions. After curing a total of two weeks, pan-
els were cut into test samples using an 8-inch diamond-blade con-
crete saw as shown in Figure 17c. The nominal dimensions of all
test samples are shown in Table 1.
Nominal thickness of a-ll ferro-cement samples was 1/2 inch, al-
though actual thickness varied from about 0.470- to 0.600-inch.
Variations in thickness were measured and the minimum thickness
was assumed to control the load capacity of all test panels.
Weidler, J. B. Jr., "Strength Characteristics of Mortar Containing
Dispersed Fibrous Reinforcement," M.S. thesis, Rice University
Library, Houston, Texas, April 1961.
Works, R. E., "The Effect of the Addition of Short Lengths of Steel
Wire on the Strength and Deformation of Concrete," M.S. thesis, Iowa
State University of Science and Technology, Ames, Iowa, 1964.
Williamson, G. R., "Use of Fibrous Reinforced Concrete in Structures
Exposed to Explosive Hazards," Ohio River Division Laboratories, U.S.
Army Corps of Engineers, Cincinnati, Ohio, August 1965.
JOHN A.' BLUME &8 ASSOCIATES. ENGINEERS
2. Fiber Reinforced Concrete - continued
Birkimer, D. L., "Fibrous Concrete Under Dynamic Tension," M.S.
thesis, University of Cincinnati, Cincinnati, Ohio 1965.
Wines, J. S., and G. C. Hoff, "Laboratory Investigation of Plastic-
Glass Fiber Reinforcement for Reinforced and Prestressed Concrete,"
Report 1, Miscl. Paper No. 6-779, p. 62, U. S. Army Engineer Water-
ways Experiment Station, Vicksburg, Mississippi, February 1966.
Shah, S. P., "Non-Linear Behavior and Composite Nature of Concrete
and Fiber Reinforced Concrete," Proceedings of the ASCE Joint
Specialty Conference on Optimization and Nonlinear Problems, Chicago,
April 1968.
Goldfein, S., "Fibrous Reinforcement for Portland Cement," Modern
Plastics, Vol. 42, No. 8, April 1965.
Agbin, C. C., "Concrete Reinforced with Glass Fibers," Magazine of
Concrete Research, Vol. 16, No. 49, December 1964.
Vasilos, T., and B. G. Wolff, "Strength Properties of Fiber-Rein-
forced Composites," Journal of Metals, May 1966.
Majumdar, A. J., and J. F. Ryder, "Glass Fiber Reinforcement of
Cement Products," Glass Technology, Vol. 9, No. 3, June 1968.
Grimer, F. J., and M. A. Ali, "The Strengths of Cements Reinforced
with Glass Fibers," Magazine of Concrete Research, Vol. 21, No. 66,
March 1969.
JOHN A. BLUME & ASSOCIATES, ENGINEERS
2. Fiber Reinforced Concrete - continued
Williamson, G. R., "Fibrous Reinforcement for Portland Cement Con-
crete," Tech. Report #2-40, U. S. Army Eng. Div. Corps of Eng., Ohio
River Div. Lab., May 1965.
McKenny, J. L., "Tensile Strength of Steel Fiber Reinforced Con-
crete," Department of Civil Engineering, M. S. thesis, Clarkson
College of Technology, May 1964.
Williamson, G. R., "Response of Fibrous, Reinforced Concrete to
Explosive Loading," Tech. Report #2-48, U. S. Army Eng. Div. Corps
of Eng., Ohio River Div. Lab., January 1966.
Irwin, G. R., "Analysis of Stresses and Strains Near the End of a
Crack Transversing a Plate," Journal of Applied Mechanics, Vol. 24,
p. 361, 1957.
Shah, S. P., and B. V. Rangan, "Fiber Reinforced Concrete Properties,"
ACI Journal, Proceedings, Vol. 68, No. 2, February 1971.
Mather, B., and R. V. Tye, "Plastic - Glass Fiber Reinforcement for
Reinforced and Prestressed Concrete; Summary of Information available
as of July 1, 1955," Report No. 1, Technical Memorandum No. 6-421,
p. 57, U. S. Army Corps of Engineers, Waterways Expt. Station, Vicks-
burg, Miss., November 1955.
Kaplan, M. F., "Strains and Stresses of Concrete at Initiation of
Cracking and near Failure," Proceedings, A.C.I., Vol. 60, No. 7,
July 1963.
JOHN A. BLUME & ASSOCIATES, ENGINEERS
APPENDIX A
PRELIMINARY CRITERIA FOR DESIGN OF FERRO-CEMENT SHELLS
123
APPENDIX A
PRELIMINARY CRITERIA FOR DESIGN OF FERRO-CEMENT SHELLS
The following preliminary criteria for design of ferro-cement structures
are based on an evaluation of a state-of-the-art survey, current design
and construction practices, and the results of a limited testing pro-
gram. These criteria are specifically related to wind tunnel
structures such as the proposed large-scale, subsonic wind tunnel
Drive Section. More extensive testing as discussed in Chapter IX
of this report will be required to establish final criteria for structural
use of ferro-cement.
A. Loads
1. Vertical Loads:
a. Dead load of ferro-cement shell and ribs.
b. Floor live load (Uniform Building Code, 1970 Edition,Table No. 23-A).
c. Roof live load (Uniform Building Code, 1970 Edition,Table No. 23-B) -- 20 psf basic.
2. Lateral Loads:
a. External wind loads in accordance with UBC (Section 2308).
b. Seismic loads should be in accordance with the appropriatespectral acceleration curves corresponding to the maximumprobable earthquake that should be determined for the site.The dynamic response of ferro-cement structures to seismicmotion should be considered in final design. This state-of-the-art procedure is now used in the design of majorfacilities throughout the United States such as nuclearpower plants and large office buildings. It provides abetter picture of the response of structures and equipmentto possible earthquake motions and hence enables theengineer to more efficiently design the facilities toresist seismic motions. Structural elements should bedesigned with increased allowable stresses for resistingthe maximum seismic loading. It is recommended that adetailed study be made to develop final seismic designcriteria before construction designs are initiated.
JOHN A. BLUME & ASSOCIATES, ENGINEERS
3. Operational loads as required for specific structure. A
typical operational pressure diagram based on recommendations
of NASA Ames Research Center personnel for a large-scale, sub-
sonic wind tunnel is shown in Figure A-1.
B. Design Standards
1. Mortar: Mix design considerations shall be based on current
state-of-the-art for ferro-cement and applicable specifications
of the ACI Code (ACI-318-72). Acceptable standards for mortar
mix materials and design are:
a. Aggregate shall be in accordance with ASTM C 33 for fineaggregate or ASTM C 330 for lightweight fine aggregate.In general, aggregates must be of the highest qualityand free of deleterious substances. When tested inaccordance with ASTM C 40, the material should beessentially free of organic impurities. Grading ofaggregate should be such that 100% passes a No. 8sieve and should, in general, provide a mortar withhigh density and good workability.
b. Cement shall conform to ASTM C 150.
c. Grading of aggregate and ratio of aggregate-to-cementshall be carefully controlled to provide uniform propertiesthroughout the structure. Control of fineness modulusshall conform to ASTM C 33 or C 330. Similarly, water-to-cement ratio shall be carefully controlled. Thespecified mix proportions should be maintained throughcareful control of batching operations and unit weight ofmortar. Compression tests at 1-, 7-and 28-day curingperiods shall be taken to check uniformity.
d. Admixtures used shall conform to Federal SpecificationSS-P-570b and ASTM Standards C 260, C 494 or C 618.Entrapped air should be measured and kept to a minimum.The use of additives not covered by the above specifica-tions, such as polymer latex additives, shall be basedon test data to verify compliance with specified mortarcharacteristics.
e. Workability of mortar, as established by grading, cementcontent, water content and additives, shall be consistentwith obtaining the highest quality ferro-cement construction.
-/oHN LA. BLUME & ASSOCIATES, ENGINEERS
R:-VDtYe HERVrO -NAi -
" 1CLOSED C/ICUI 7 " WI/JD
TUM7 EL PLAN
c DE K L o0
126
_FG UR E A- - Y TPICAL P/RESSU/Re o/ JR/AUT1/oN,FOR LAReG.E-SCALCe ,SUSOM/C WINO TUNVAeL
JOHN A. BLUME & ASSOCIATES, ENGINEERS
20oZ40
foo
4)
0l
4o.
t?
Q.q
80
eo
f. Aggregate selection and mix design should minimizedrying shrinkage, which should be less than 0.03 percentfor laboratory control samples, as determined byASTM C 341.
2. Reinforcing Steel:
a. Wire mesh shall be welded wire mesh or woven wire meshconforming to ASTM A 185. Wire shall conform toASTM A 82.
b. Reinforcing bars shall be as specified by ASTM A 615 orASTM A 616 for Grade 60.
3. Surface Cracking: Extent of surface cracking and maximum crack
width allowed shall be as permitted for specific structural
application.
4. Ferro-cement Design Stresses: Design stresses should be based
on consideration of the material as an equivalent homogeneous
cross section. Design values are dependent on the type and
configuration of materials (mortar and reinforcing) used and
are defined by the current research results representing the
state-of-the-art in ferro-cement, which are discussed in
Chapter III of this report. Design stress for a specific
cross section should be as follows, where F.S. represents an
appropriate factor of safety:
a. No cracking allowed:
Design stress = First cracking stress
F.S.
b. Cracking permitted:
Ultimate stressDesign stress = Ultimate stress
F.S.
Design stress as related to allowable -crackwidth is currently under study. SeeChapter III of this report.
The value of F.S. is based on the intended structural use and
type of loading, ferro-cement research results and the principles
A-3127JOHN A. BLUME & ASSOCIATES, ENGINEERS
of reinforced concrete design. For simplicity, a factor of
safety was used for determining design stress. It should be
noted, however, for reinforced concrete design the value of
design stress is based on statistical considerations and on
keeping within a maximum probability (generally extremely
small) of failure.
5. Structural Support: Supports and connections for ferro-cement
structures shall be capable of resisting all loads on the
structure. These shall conform to current design practice and
be verified by load tests.
C. Fabrication Standards
1. Dimensional Tolerance: Fabrication and erection procedures
shall ensure structural tolerances as required.
2. Surface Quality: Surface texture shall be as required in
finished structure. Design and fabrication methods shall
ensure adequate surface quality for a specified service life.
3. Durability and Corrosion Resistance: Ferro-cement design and
fabrication shall ensure durability of the material. Careful
consideration shall be given to mortar corrosion resistance
and permeability, reinforcement, and placing and curing methods.
Galvanized reinforcement shall be used whenever possible.
Adequacy of the design or the need for protective coatings
shall be established by appropriate tests for each structure.
JOcHN A. LUME & ASSOCIATES. ENGINEERS
APPENDIX B
REVIEW OF FEASIBILITY OF USING FERRO-CEMENT CONSTRUCTION
FOR PROPOSED NASA WIND TUNNEL DRIVE SECTION
129
APPENDIX B
REVIEW OF FEASIBILITY OF USING FERRO-CEMENT
CONSTRUCTION FOR PROPOSED NASA WIND TUNNEL
DRIVE SECTION
The following report prepared by David J. Seymour of David J. Seymour,
Naval Architects and Marine Consultants, contains a feasibility review
and cost estimate for the proposed use of ferro-cement in wind tunnel
construction. This work is based on Mr. Seymour's experience with the
design and construction of a large ferro-cement cargo barge which is
discussed in the preceding text.
The wind tunnel structure on which the ferro-cement cost estimate in
this appendix is based is the Drive Section shown in Figures 22 and
23 of the preceding text. This structure contains 20 drive units
(4 high by 5 wide) and is 200 feet long. The ferro-cement portions
are the shrouds and nacelles which provide aerodynamic surfaces for
the flow of air past the drive unit fans. Total surface area of
these ferro-cement shrouds and nacelles is approximately 900,000 square
feet. The cost estimate developed by Mr. Seymour is given in terms of
both total cost for 900,000 square feet and unit cost on a square foot
basis. Using the unit cost, estimates can be made of the ferro-cement
costs for similar structures.
VB °JOHN A. BLUME & ASSOCIATES. ENGINEERS
I
David J. SeymourNAVAL ARCHITECTS - MARINE CONSULTANTS
Word Trade Center - Suite 330 Board of Trade Tower - Suite 1012Embarcadero at Market 1177 West Haitings StreetSan Francisco, Ca 94111 Vancouver 1. B.C. CanadaTelephone (415) 398-8454 /398-8415 Telephone (604) 685-8394
April 13, 1972File 212
John A. Blume & Associates, EngineersSheraton Palace Hotel100 Jessie StreetSan Francisco, California 94105
Attention: Mr. Roland L. Sharpe, Executive Vice President
SUBJECT: FEASIBILITY REVIEW OF FERRO-CEMENT FORPROPOSED NASA WIND TUNNEL
Encl. 1) DJS SK. 212 - Ferro-Cement De.sign for WindTunnel
Gentlemen:
In accordance with your request, I have completed a generalfeasibility review of the utilization of ferro-cementpanels for lining of the surface areas of shrouds andnacelles in the drive section of subject wind tunnel.
The objectives of my review were primarily to:
a) Consider general design, fabrication and assemblymethods of ferro-cement for this application.
b) Estimate unit costs based on designs of item a)above.
c) Determine the. engineering feasibility of employingferro-cement in relation to the current "state ofthe art" for this material.
1. DESIGN, FABRICATION & ASSEMBLY METHODS
a) Data and design criteria used in this review:(as given by Blume Engineers)
No. of SHROUD/NACELLE UNITS - up to 20 (4 high -5 across).
Height to top shroud above ground - about 210 ft.
Air Pressure Loading - Shrouds 100 PSF, Nacelles0 PSF.
Surface Tolerances - Offsets from Fan
0 - 20 ft. = 0 in.
20 - 30 ft. = +3/32 in.
Over 40 ft. = + 1/2 in.
Steel Supporting Structure - in place for shroudswith support points,at 20 ft. intervals.
b) General Design Considerations
The design criteria for present ferro-cementconstruction, primarily in marine'application,have been based on strengths to meet hydrostaticloading (up to 700 PSF), stresses due to hoggingand sagging bending moments, water tightness,impact damage, fire and corrosion resistance.These are not present in the requirements for sub-ject wind tunnel. However, two new design para-meters have been added, namely fan induced vi-bration loading and wind erosion. Little data isavailable on these factors for ferro-cement and,although not considered a major problem area, itis recommended that some R&D effort be directedto determine their effects.
Due to repetitive compound shapes involved, re-quirements for accuracy in surface tolerances and,inaccessiblilty for efficient "in place" fabri-cation, the precast method is the obvious solution.Precasting would permit accurate surface shape control
J.A. Blume & Associates
(side against mold), precision forming of jointedges and, fabrication under quality controlconditions.
Optimum panel size should be based on unit weightand costs, to provide suitable panel strength forself-supporting, ease of transportability to site,and for efficient assembly operations.
The latter will most probably control panel sizeand weight because of the height of shrouds aboveground and the interference of pre-installedsteel structure.
c) Review of Precast Method
Shroud panels should be easily fabricated byemploying a male mold representing the full lengthof 1/2 of the diameter of a shroud. Panels shouldbe cast in 20 ft. lengths to match structure supportpoints and be 1/8 or 1/4 circle sections. Precastribs and stiffening member would be cast intopanel when latter is poured. See Sketch Encl. 1).
Nacelle panels should similarly lend themselves toconstruction but on a female mold. Due to their20 ft. diameter, it appears feasible that internalframing of ferro-cement material could reduceconsiderable amount of nacelle steel supportstructure. Consideration might be given to eli-minating all nacelle steel structure by introducinga prestress (post-tension) system to accommodatethe spans between struts. See Sketch Encl. 1).
In employing the precasting method, only one moldfor shrouds and one for nacelles would be required..Accurate surface dimension control of the compoundcurves would be insured being faced on the mold.Finishing and application of coatings would be done
. prior to assembling panels in'place'.
d) Ferro-Cement Panel Design
Although a wide variety of lay-up materials, con-figurations and cement mortar may be suitable for
133
3 - April 13, 1972
J.A.' Blume & Associates-
this application, the writer has selected a paneldesign based on the most advanced developments inmarine design. Considerabl.e research data andactual construction, including approval by marineregulatory bodies, are incorporated in its designso that .it.should be a sound basis for cost andfeasibility evaluation.
Principal Cheracteristics of Panel
Thickness - 5/8" overall
Lay-Up - 1 layer WWM' 1/4"
1 layer Rod 1/4"
2 layer WWM 1/4"
x 1/4" .x 21 ga.(S 80,000psi)
dia. on 2" centers(S 80,000 psi)
x 1/4" x ~1 ga. (Sy 80,000psi)
Stiffner Spacing - 4 ft.
Cement Mortar - Lightweight Cement, Crushed andUncrushed Saturnlite Sand, Pozzolanand Pozzolith and Special Additives.Slump - 0
Vibrators - extensive use of vibrators on mold andhand type to insure full penetration
Curing - accelerated at low temperature
See Sketch Encl. 1)
e) Assembly Methods
In the writer's opinion, the assembly method willbe a major factor affecting the feasibility ofusing ferro-cement for this project.
The controlling factors for optimum panel size andweight would be for ease of handling, positioning,alignment and securing. Also unit cost per sq. ft.and joint length would be reflected in optimizing.
The following method was selected for analysis:(assume nacelle steel work to be installed aftershroud panels are in place)
134
- 4 - April'13, 1972
J.A. Blume & Associates -5- April 13, 1972
i) Transfer 20 ft. long precast and prefinishedshroud panels by hoisting to shroud level.
ii) Transfer panel into position within shroudby special dolley, track and jig.
iii). Set panel in place and align with Laser.
iv) Epoxy, grind and finish joints - touch upas required.
v) Install steel struts and nacelle steel framing.
vi) Install 20 ft. nacelle paneks, similar to stepsfor shroud.
2. COST ESTIMATE
The following assumptions were included in preparingcost estimate.
Total Ferro-Cement Panel Area - approx. 900,000 sq. ft.(includes area of concrete drive unit support -20 units)
Allowance for R&D, Design and Engineering.
Material and Labor for Molds.
Material and Labor for Manufacturing.
Material, Equipment and Labor for Assembly - hoists,jigs, dollys, staging, etc.
Contingencies (margin, changes, escalation, etc.) of 15%.
SUMMARY
A. R&D, Design & Engrg.
B. Molds
C. Manufacturing
D. Assembly
E., Contingen'cies 15%
TOTAL
$ TOTAL
150,000
215,000
4,050,000
600,0005,015,000
752,000
5,767,000
UNIT$/Sq.Ft.
0.17
0.24
4.50
0.67
0.4836.41
135
J.A. Blume & Associates
3. FEASIBILITY
In the writer's opinion, the concept of employingferro-cement material for- lining of the shroud andnacelle surfaces is feasible from both a constructionand economic viewpoint.
The proposed application here of ferro-cement is wellwithin the current "state of the art" of the material.This is further supported by the fact that it is ex-posed to less severe environment and loads than thosefound in marine practice.
Also rapid developnen.ts are presently unrder'wdy, bycommercial and governmental agencies, in the researchand application of this material to improve both itsstrength and weight characteristics. These improve-ments no doubt will be available to incorporate in thisproject.
In addition, due to the large areas of ferro-cementinvolved, efforts can be afforded in optimization ofdesign, fabrication and assembly methods to produceimprovements over the writer's assumptions used inthis evaluation.
fiery truly yours,
\ \(
DAVID J, SEYMOUR
DJS/rb
136
- 6 - April 13, 1972
David J. SeymAVAL ARCHITECTS-MARINE COHWORLD TRADE CENTER - UI
EMBARCADERO AT MARIAN FRANCISCO.e A"I-'O
IIi
I.. t.
hQvnie wfl6WwS
PA-t.L s c qus
S'V4D e IbpU* I 96 to ttlC~·AC4 e \ZdO n = 3'0o 8
W~JO 1C~~VN 4~|SS~~ GYO CEVE Wt DE54 iu*lhuZ
3Is SK, -7.i37
APPENDIX C
LABORATORY REPORT ON STRUCTURAL INVESTIGATION OF
FERRO-CEMENT SPECIMENS
138
SAN JOSE STATE COLLEGE125 South Seventh Street, San lose, Caliiornia 95114
SCHOOL OF ENGINEERING (408) 277 - 2488Department of Civil Engineeringand Applied Mechanics
Mr. Roland L. Sharpe
Executive Vice President
John'A. Blume & Associates, Engineers
100 Jessie Street
San Francisco, California 94105
Dear Mr. Sharpe:
Enclosed are three copies of the Laboratory
Report on Structural Investigation of Ferro-Cement
Specimens.
Please inform me if you wish to discuss
the report in more detail. It has been a pleasure
to be of service to your firm.
Very truly yours,
W. J. Ven ti, Ph. D.
Professor of Civil Engineering
139
LABORATORY REPORT
on
STRUCTURAL INVESTIGATION OF
FERRO-CEMENT SPECIMENS
for
John A. Blume & Associates, Engineers
100 Jessie Street
San Francisco, California 94105
by
William J. Venuti, Ph. D.Department of Civil Engineering
and
Applied Mechanics
School of EngineeringCalifornia State University, San Jose
San Jose, California 95114
June 1972
140
STRUCTURAL INVESTIGATION OF
FERRO-CEMENT SPECIMENS
Introduction
John A. Blume & Associates of San Francisco,
California, has engaged in a preliminary study of the
static, dynamic, and fatigue characteristics of ferro-
cement as a structural material. The Department of Civil
Engineering and Applied Mechanics of the California State
University, San Jose, has collaborated with John A. Blume
& Associates in conducting the experimental phase of the study.
This report presents an account of the experimental
design and description of the laboratory equipment utilized
in conducting the experimental study. A description of the
ferro-cement mix design and reinforcement, laboratory data,
and analysis of the data is presented separately by John A.
Blume & Associates.
Laboratory Tests
The following tests were conducted on the ferro-
cement panels and specimens:
141
1. Flexural static tests
2. Flexural fatigue tests
3. Flexural vibration tests
4. Beam shear tests
5. Cube compressive tests
6. Slab compressive' tests
7. Wire mesh strand tensile tests
8. Slab tensile tests
9. Air Flow tests
The following sections describe the specimens tested,
laboratory equipment, instrumentation and method of loading
of the various tests.
The entire laboratory study was conducted in the
Advanced Structures Laboratory of the California State Uni-
versity, San Jose.
All load indicator systems used in this experimental
program are calibrated and certified to be traceable to the
U.S. Bureau of Standards.
FLEXURAL STATIC TESTS
Specimen Preparation
Each flexural specimen was measured for thickness
with a 0.001 inch accuracy micrometer. The measurements were
taken at the corners of the zone of maximum bending moment.
-2-14Z
At the bottom surface of the end of each flexural
panel, quick setting mortar (Hydrocal) was placed in the vi-
cinity of the support area. With the Hydbrcal in a plastic
state, the panel was slightly pressed against a rigid level
table to obtain smooth and parallel surfaces of the beam end
supports.
At the two loading points on the top surface of the
beam, additional transverse strips of Hydrocal were placed.
Prior to the setting of the Hydrocal, 1 inch wide by i inch
thick teflon pads were placed at the loading points. The
pads having a length equal to the width of the beam, were
pressed into the plastic Hydrocal with the channel loading
device. This procedure assured uniform bearing of the loading
device during testing.
Equipment and Instrumentation
The purpose of this series of tests was to determine
the load-deflection relationship, cracking load, and ultimate
load of each flexurao specimen.
A 12,000 lb. capacity mechanical type Tinius-Olsen
universal testing machine was used to apply the load. The
120 lb. loading range was used to obtain the load-deflection
curve for the initial phase of each test. The 600 lb. loading
range was used to carry the test to ultimate loading.
-3-
143
The beam end supports were free to rotate to elim-
inate longitudinal restraint of the beam. The load was applied
at two points symetrically located about the beam centerline.
The load was applied by means of an aluminum channel which
was accurately placed over the teflon strips. A spherical
loading head attached to the upper platen of the testing machine
applied the load to the loading channel.
Deflections were measured at the beam centerline with
a Tinius-Olsen Model D-2 Deflectometer which incorporates a
linear voltage differential transformer (LVDT). The deflection
of the beam was magnified lOOX and recorded on the machine
12 inch wide continuous chart recorder. The load-deflection
curve was directly displayed on the recorder paper.
The loading rate for each test was set at 0.2 inch
per minute.
In order to determine the load of the formation of
the first flexural crack, silver conductive paint (with a
butyl acetate base) was applied to the bottom surface near
each outer edge along the length of the beam in the zone of
maximum flexure. An electrical circuit was completed with a
3 volt ohmeter. Upon the formation of the first flexural
crack, the circuit breaks and the load at first crack is
recorded.
-4-
144
FLEXURAL FATIGUE TESTS
Specimen Preparation
The specimens that were subjected to repeated load-
ing in this series of tests were prepared in the same manner
as those specimens used for the flexural static tests.
Equipment and Instrumentation
The purpose of this series of tests was to determine
the number of cycles of repeated loading of specified magni-
tudes of load required to produce the formation of a flexural
crack.
The same support and loading arrangement as used in
the flexural static tests was used for these tests. However,
the entire loading apparatus was contained in the loading
frame of a 120,000 lb. Baldwin universal testing machine. The
machine was used only for the purpose of positioning the load-
ing head and specimens in a vertical direction.
The repeated loading was applied to the specimen by
means of a hydraulic closed loop servo system. A 3000 pound
capacity MTS hydraulic ram under load control was supported by
the upper head of the testing machine.
The ram was actuated with an MTS servo-controller
which received signals from a double bridge 500 pound capacity
-5-
145
electrical strain gage load cell. Hydraulic power supply
was provided by a 10 gpm pumping unit. A micro-switch was
placed beneath the ferro-cement specimen for the purpose of
stopping the power supply in the event of panel failure during
the application of the repeated loading. An LVDT was also
placed beneath the center of the beam specimen to monitor
centerline displacement at various times.
A Model 126B VCF/Sweep MTS Function Generator was
used to supply the sinusoidal load input. The loading rate
was maintained at 12 Hz. for all tests.
The magnitude of load and displacement were displayed
on a Type 564 Textronix Storage dual beam oscilloscope with
a Type 3C66 Carrier Amplifier and a Type 2B67 Time Base. The
oscilloscope signals were channeled through a Model 297
Sanborn strip chart recorder.
To determine the number of cycles of repeated loading
at which the first flexural crack occurred, a relay system
was installed. The conductive paint circuit on the panel was
connected to a relay which was placed in the circuit of an
electric timer. The system was designed to break the circuit
of the electric timer upon the formation of a gap in the paint
circuit (caused by a flexural crack in the test panel). The
number of cycles at first crack was determined by obtaining
the product of the loading frequency and elapsed time.
-6-
146
FLEXURAL VIBRATION TESTS
Specimen Preparation
The three specimens that were tested for the pur-
pose of obtaining the dynamic characteristics of the ferro-
cement panels were prepared with Hydrocal at the support points
and loading points similarly to those specimens used for the
flexural static tests.
Equipment and Instrumentation
The purpose of this series of tests was to determine
the fundamental flexural frequency of vibration and the damp-
ing coefficient for ferro-cement panels in a cracked and un-
cracked condition.
The same support and loading arrangement as previously
described was used for this part of the testing program.
The testing procedure was as follows. The test panel
was loaded downward statically to produce a positive moment
corresponding to a predetermined load. The force of the load-
ing ram was transmitted to the loading device on the test
panel by means of a 12 inch length of i inch diameter steel
rod. The steel rod was then abruptly pulled away from the
loading ram and loading device in order to excite the panel
to vibrate at its natural frequency. In some cases, fixed
-7-
147
weights were placed on the loading device to prevent the panel
from vibrating away from the end supports. In one case, weights
were placed over the supports to assure that the beam main-
tained contact with the supports during vibration.
An LVDT placed beneath the centerline of the test
panel was used to obtain the displacement variation with time.
The displacement-time response was displayed on the oscillo-
scope with a persistent image of the beam. A photograph of
the oscilloscope screen was taken with a Polaroid Land Oscil-
loscope camera.
The horizontal time rate of the oscilloscope beam was
set at 0.2 sec. per cm. or 2 seconds for a full screen sweep
of 10 cm.
The vertical scale of the beam was set at rates of
.05 Volts, 0.1 Volts, 0.2 Volts, and 0.5 Volts/cm. The rela-
tionship between beam deflection and oscilloscope beam move-
ment was established prior to the vibration tests.
BEAM SHEAR TESTS
The supports and loading points were prepared with
Hydrocal as previously discussed. The load was applied in a
hydraulic universal testing machine at a rate of 0.05 in./min.
-8-
148
CUBE COMPRESSIVE TESTS
The two-inch cubes of cement mortar were loaded in
a hydraulic universal testing machine at a rate of 2000 psi/min.
A spherical loading block was used to apply the load.
SLAB COMPRESSIVE TESTS
The loaded edges of the slab compression specimens
were prepared with Hydrocal to assure uniform loading. A
spherical loading block applied the load at a rate of 2000 psi/
min. in a hydraulic universal testing machine.
WIRE MESH STRAND TENSILE TEST
Each length of wire was tested in tension in a hy-
draulic type universal testing machine. Each end of the wire
was gripped in flat face wedge-type grips over a length of
3 inches. The free length of wire under tension was 12 inches.
The rate of loading was approximately 50 lb./minute.
SLAB TENSILE TESTS
Each tensile specimen was prepared by applying a
3S inch length of Hydrocal to each face at each end. The
-9-
149
Hydrocal was placed in such a manner that the surfaces
were smooth and parallel to each other. Conductive paint
strips were vertically placed along each face near each ver-
tical edge to determine the tensile load at first crack.
The specimens were subjected to a tensile force in a
mechanical type universal testing machine. The flat face
wedge type grips were shimmed to eliminate bending of the spe-
cimen under loading. The tensile force was applied at a rate
of approximately 500 pounds per minute.
AIR FLOW TESTS
The purpose of these tests was to observe and
examine potential structural deterioration resulting from a
constant stream of air trained on the surface of a ferro-cement
panel at a pressure of 6 psi. for a period of 7 days.
In this study, two panel surfaces were subjected to
an air stream. One surface was the side of the panel that
was adjacent to the form during construction and the other sur-
face was one which was trowelled during the finishing operation.
An uncracked section of a 24 inch long beam was used
for each test. The panel was placed in an upright position
and fixed at an angle of 45 degrees to the air stream. A
2 in. by 2 in. square area of the panel surface was designated
-10-
150
as the test zone and placed so its center coincided with the
center of the air stream.
The calibration of the apparatus was made as follows.
One end of a ½ inch diam. air hose was inserted into a hole
in a 3/4 in. thick wooden board and the opposite end of the
hose was connected to a pressure gage. The board was then
subjected to a steady jet of air which was issued from the
valve of an 80 psi air supply line. The wooden board was
positioned so that its surface was normal to the direction of
the air stream. The opening in the panel where the hose was
installed was centered on the air stream while pressure read-
ings were taken. Based on readings of several trials, a pres-
sure of 6 psi was obtained when the panel was placed at 3½
inches from the face of the valve.
Using this relationship as a basis, the ferro-cement
panels were also placed at a distance of 3½ inches to obtain
a pressure of 6 psi, The panels were rotated at 45 degrees
to the air stream with the center of the test zone remaining
at 3½ inches from the valve.
The two panels were tested concurrently at different
locations.
-11-
151
APPENDIX D
REVISION OF DESIGN STUDY OF POWER SECTION
FOR PROPOSED V/STOL WIND TUNNEL
isZ
APPENDIX D
REVISION OF DESIGN STUDY OF POWER SECTION FOR
PROPOSED V/STOL WIND TUNNEL
The cost estimate for ferro-cement shrouds and nacelles, which was
presented in Chapter VI and Appendix B of this report, was used to
revise the cost estimate for Concept III (Ferro-cement) for the Drive
(Power) Section that was the subject of the March 22, 1971, John A.
Blume & Associates, Engineers report "Conceptual Design Study of
Power Section for Proposed V/STOL Wind Tunnel."
A summary of revised cost figures for this Power Section is contained
in Table D-1. This summary contains cost figures for the entire struc-
ture including structural steel framing, painting, piles, and concrete,
including ferro-cement shells for shrouds and nacelles. The cost
figures in Table D-I reflect the following basic revisions to the
March 22, 1971, estimate:
1. Ferro-cement shrouds and nacelles are 5/8-inch shells with concrete
stiffening ribs.
2. Ferro-cement is used throughout for shrouds and nacelles; no
structural steel plate is used.
3. Slight increase in structural steel framing results from increased
weight of revised ferro-cement shell and concrete ribs.
The unit price of $6.41 per square foot for a 5/8-inch stiffened ferro-
cement shell is based on the cost estimate in Appendix B. It should
be noted that this unit price was developed for the revised Drive
Section shown in Figures 22 and 23 of this report, which is 200 feet
long with 20 drive units (4 high by 5 wide) containing approximately
JOHN A. BLUME & ASSOCIATES, ENGINEERS
TABLE D-I
REVISED COSTS FOR POWER SECTION STRUCTURE
( FERRO-CEMENT)
(Costs for Power Section 244 feet long containing 18 drive units,described in the March 22, 1971, Blume report.)
Number of Cost forDescription Unit Unit Cost Units Total Units
Ferro-cementshrouds SF 6.41* 773,000 4,955,000nacelles (including
struts) SF 6.41* 468,000 3,000,000
Structural Steel
Shapes-Nacelles T 1,000.00 3,680 3,680,000
Shapes-Shrouds T 1,000.00 5,789 5,789,000
Shapes-Motor supports T 1,000.00 771 771,000
Painting
Shapes SF 0.15 1,430,000 21"5,000
TOTAL COST $26,712,000
*Ferro-cement unit costs are based on cost estimate inAppendix B which is for a revised Power Section 200 feetlong with 20 drive units and approximately 900,000 squarefeet of ferro-cement.
JOHN A. BLUME & ASSOCIATES, ENGINEERS
/c;4A01
900,000 square feet of ferro-cement. Al ughth Power Section dis-
cussed in the March 22, 1971, Blume report (and revised in Table D-l)
is 244 feet long with 18 drive units and 1,241,000 square feet of
;ferro-cement, the loading conditions and construction problems are the
-same. Therefore, the unit cost of $6.41 is applicable to both struc-
/,/tures.
The revised costs in Table D-1 show an increase for the ferro-cement
relative to the March 22, 1971, estimate. This increase reflects the
more in-depth feasibility and cost studies contained in the present
report and the revised construction recommendations. Based on con-
clusions reached in this report, the ferro-cement unit cost used in
Table D-1 is conservative since advances in ferro-cement material re-
search and design procedures and especially improvements in construc-
tion methods making greater use of automated fabrication techniques
should lead to reductions in the estimated costs. Any reduction in
fabrication cost will result in significant overall ferro-cement cost
reduction because fabrication comprises over two-thirds of the total
